BAUDELAIRE AND STONE TECHNIQUES FOR RASTER GRAPHICS SIGGRAPH '87 TUTORIAL COURSE NOTES DOCUMENTATION GRAPHICS === length: 1 in Techniques for Interactive Raster Graphics Techniques for Interactive Raster Graphics Patrick Baudelaire* and Maureen Stone Xerox Palo Alto Research Center ABSTRACT: The visual quality of raster images makes them attractive for applications such as business graphics and document illustration. Such applications are most fully served using interactive systems to describe curves, areas and text which can be rendered at high resolution for the final copy. However, to present such imagery in an interactive environment for moderate cost is difficult. Techniques are presented that provide solutions to the problems of scan conversion, screen update, and hit testing for a class of interactive sytems called illustrators. The design rests on the use of software display file encoding techniques. These ideas have been used in the implementation of several illustration programs on a personal minicomputer. Key Words: computer graphics, interactive graphics, display encoding, chain encoding, run-length encoding, scan conversion, illustration systems. *Present Address: TECSI, 29, rue des Pyramides, 75001 PARIS, FRANCE Introduction Pictures are a part of a large number of applications. A certain class of pictures can be referred to as illustrations. That is, the point of the picture is to illustrate a principle as part of an article or presentation. Good illustrations are visually interesting, but are judged more on their content than on their artistic merit. With respect to picture complexity, illustrations are simple in that there are a moderate number of shapes, clearly bounded by curves and lines. The image is essentially two dimensional, and color is often restricted to flat filled areas; that is, uniform colors and simple textures. Illustrations are important in business and publishing environments today. The advantages to be gained from using a computer to generate these images are similar to the ones for word processing: images are easily modifiable, pictures can be filed and copies easily generated, subimages can be libraried to facilitate the composition of a series of illustrations. However, to be acceptable outside of the experimental environment, the image quality of illustrations must be high. Curves and lines must be smooth, text must be represented in a variety of fonts, objects must be accurately positioned. While each illustration may contain a limited number of colors, a wide range must be available. Given the type of imagery desired, a raster display will give much better representation of the picture than a line oriented display. The constraint of high image quality, even after scaling and repositioning operations, means that the picture must be stored using a high precision representation. Furthermore, if the data base has sufficient precision, the set of affine transformations can be used as tools for generating images. The process of designing an image is the mapping of some visualization onto a medium. Therefore, that any effective tool should provide visual feed-back as the image was formed. Furthermore, the image should be built up in a two-dimensional manner, that is, by pointing to positions on a page, which indicates an interactive system. In this paper, such a system will be called an illustrator. Given a high precision data base, such as endpoints for lines, coefficients for splines, plus information about the resolution of the display, there are many standard techniques for displaying straight lines, curves and areas (5). This process is known as scan conversion. There are a number of problems in using these techniques in an illustrator, specifically: f All straight lines and curves must be represented with a finite width. In an illustration, using lines of different widths is part of the visual effect. Therefore, displaying these shapes involves more than just digitizing the curve. f Areas are discribed by their outline. Therefore, it is necessary to compute which points are inside the outline and color them in accordingly. While there are many techniques which address this problem (10, 11) in an interactive system the speed at which this can be done is an important factor as the display is constantly changing as the user works. f In an interactive system, the display is constantly changing. For speed considerations, it is desirable to do this incrementally, updating only the areas that are affected by the change. Incremental refresh can cause problems displaying the correct overlap order for objects. In an interactive system it is necessary for the user to select objects to be manipulated. A natural way to do that is to point to the object. Therefore, given a point on the display, it is necessary to determine which object has been selected. This is called hit testing. Hit testing is the inverse of the display problem, and is even more speed critical than refresh. This paper will discribe methods for solving these problems that achieve a reasonable compromise in simplicity of implementation and response time. The solution is based on using display encoding techniques which provide a representation of the image that is compact, structured, and simple to manipulate. These techniques have been implemented in systems which have been used to successfully create illustrations. Examples are included at the end of the paper. Data Representations and System Design Graphics systems can be categorized by the type of data used to represent the picture. One type of interactive graphics systems uses raster dots or arrays of intensity samples as the unique representation of the image. These systems use a painting model to manipulate the dots directly. These systems are easy to use, and have been shown to produce effective imagery (8, 9). However, the picture is unstructured, which limits the types of manipulations that can be performed without special hardware. The resolution is tied strongly to the display resolution, which can limit image quality. In general, to be effective visually, the images from such systems require large amounts of data, either in resolution or in bits per pixel. We have chosen a more structured approach. Given a geometric representation, one can consider that geometric elements (lines, curves, areas) are all made of filled contours. This paradigm is quite useful for applications such as the production of rasters for high-resolution printing of graphics and text (1). However, this model is difficult to implement in an interactive application because using the geometric data for display and hit testing can be expensive in terms of computation time. An interactive display file can be used to provide a definition of the picture that is structured, yet can be manipulated fast enough for an interactive system. It is natural for several representations to coexist in the design of a graphics system, usually according to levels of hierarchy. At the top level may exist a representation that embodies some specific knowledge or meaning relevant to a particular application: architectural drawings, blueprints for mechanical parts, electrical circuit diagrams, etc. Here we will consider that the top level representation that we chose aims only at giving a unambiguous and complete description of high-quality business graphics illustrations. To implement these ideas, the following systems approach was used: The workbench is a moderate resolution raster display plus pointing device attached to a 16 bit mini-computer. The final output device is a high resolution raster device such as a film recorder, phototypesetter, or raster printer. Shapes are defined by their geometry: trajectories and contours; plus style informations such as line width, colors, and textures. Trajectories can be specified by one of several mathematical schemes such as splines or other knot-based approximations, circles or other conical equations. Text is unformatted, and described in terms of position and string information plus style parameters such as font, color, and orientation. The user builds display objects by pointing at fitting points and indicating fitting methods such as straight lines or curves. All numbers are represented as floating point values to provide sufficient precision. The top level description is converted to an interactive display file, where the interactive processing, refresh and hit testing, will take place. The display file is used to generate a bitmap, that is a one bit per point rectangular array of samples (13), which is used as the refresh buffer for the display. This structure is summarized in figure 1. The Interactive Display File The display file representation is used for refresh and hit testing. In designing such a representation, the first consideration is to what precision the objects should be encoded. Clearly, display resolution is sufficient for refresh. Since the user cannot specify a position by pointing which is more precise than the display resolution, it is also sufficient for hit testing. Both hit testing and incremental refresh involve scanning through the display file to determine what objects are in a specified area. Therefore, partitioning the display file to facilitate culling will increase performance. It is important that the display file be reasonably compact, yet not so difficult to generate or decode that it negates the advantages in terms of speed. It is also convenient to be able to encode all objects in a uniform manner. Given these considerations, we have applied and modified some well known encoding techniques, chain encoding of trajectories (3, 6) and run-length encoding of areas (2, 4). These techniques are essentially compressed representations of the bitmap. Chain encoding is based on the assumption that edges are continuous. Therefore, the basic representation is the differences between adjacent pixels. Run-length encoding is based on the assumption that flat areas contain most of the information in the outlines. Therefore, the basic representation is the position of the edge on each scan line. While it would be possible to run-length encode all objects, the increased structure in the chain encoding is appealing for lines. === Figure 1: Levels of representation. === The particular encoding schemes chosen permit the segmentation of each object into pieces that are independent and bounded in display size. It follows from this that the time for display of one piece is bounded too. This makes it possible to run the screen refresh as a background process. All shapes can be described by one of the two types of encoding. Thus, pointing detection can be done by a single algorithm, independent of the mathematical definition of the elements (lines, circles, conics, splines, etc.). Chain Encoding Chain encoding is a differential encoding scheme which records the screen coordinate increments between successive raster points on a trajectory. That is, from one trajectory point to the next, raster coordinates may differ only by -1, 0 or 1. Thus, a point may have eight possible successors, so each point new position could be represented in four bits. But, since common curve trajectories are monotonic for reasonably long intervals, one can take advantage of the continuity in direction to further reduce the number of bits per point. A number of schemes are possible for encoding coordinate increments. Two interesting and practical ones are described below. The first scheme uses two-bits to represent the coordinate increments (figure 2). The set of eight possible curve directions is divided into four quadrants. For each direction quadrant, the three possible coordinate increments are assigned code values 1 to 3. Code value 0 is used to indicate a change of quadrant, with the following two bits specifying the new quadrant. Therefore, the trajectory encoding is a stream of two-bit codes, starting with the quadrant number (0 to 3), followed by increment codes (1 to 3), terminated by a 0. === Figure 2: Chain Encoding, Two Bits/Direction. === The second scheme requires two streams (figure 3). The set of eight possible curve directions is divided into eight octants. Within each octant there are two possible directions. Therefore, it is possible to indicate each step within an octant with one bit. One stream contains the one-bit increment codes for a given direction octant, and the second stream contains the octant numbers and the number of steps in each octant. Besides being somewhat more efficient, it is possible to get a general idea of the behavior of a curve segment by examining the direction octants alone. An example of where this feature can be used is for defining edges for the scan converter. Using the chain encoding in the scan conversion process will be described further below. === Figure 3: Chain Encoding, One Bit/Direction. === As mentioned above, in order to gain efficiency in the screen updating and pointing detection algorithm, the encoding is fragmented into independent pieces of similar length that we call chunks (figure 4). The bounding box for each chunk can be stored to facilitate culling on refresh and hit testing. The display file contains the following information for each chunk: Screen coordinates of the starting point S Stream(s) for the chain encoding The bounding frame: H and W. It is interesting to note that if the chunk size is such that each segment contains at most N trajectory points, all these points are enclosed in a square of size 2N centered at the starting point of the segment. So even if the bounding box is not explicitly stored with the chunk, a bounding region can be computed. Run-Length Encoding Run-length encoding defines an area in terms of a starting scan line (Y) value plus a list of pairs of raster (X) values. In practice, it is more efficient to make the second X value relative to the first, so each run is defined as a starting X (SX) and a delta X (DX) value. The list of runs can be broken into chunks, such that each chunk defines a maximum of N runs. Therefore, each chunk has a bounding frame. It is desirable to make the starting X values relative to this frame, so that the chunk can be relocated simply by translating the frame boundries. (figure 5) === Figure 4: Curve Divided into Encoded Chunks. === The encoding described above works only for convex areas, specifically, it assumes one continuous run of rasters per scan line within the area. For concave shapes, there are two options: break the area into convex pieces, or introduce a flag into the list of runs that defines the number of runs per scan line. We have implemented the second option, choosing a negative starting X as a flag, signaling that the delta X value is the new number of runs per scan line. Assuming the run-length encoding describes areas at display resolutions, the starting and delta X values need be no larger than the maximum display coordinate. In practice, most runs can be described in fewer bits. Therefore, it is worthwhile to consider using a variable field for X. While maximum compression would be obtained by using a technique such as Huffman coding for field size, for simplicity we have chosen to implement two fixed fields (8 and 16 bits). The display file contains for each chunk: === Figure 5: Run-Length Encoding, Showing Chunks. === Frame boundry: upper left point S in screen coordinates (includes the starting Y value), plus frame size H and W. Field length: either bytes or words. List of run values, defined as SX relative to S, and DX relative to SX. Scan Conversion The geometry to rasters conversion scan conversion process can be decomposed in two operations: Converting the geometry into the display file representation, and converting the display file to rasters. The conversion from geometry to display file only need be performed when an object is created or the shape is changed. Because this is a relatively infrequent operation, standard techniques for digitizing curves and filling areas provide acceptable speed. The specific algorithm used for areas is given at the end of this section. Objects are displayed in back to front order to give the correct overlap information. All display, refresh, repositioning, and hit testing operations can be implemented by manipulating the display file. Display The display operation is defined as the conversion from the display file representation to the rasters which are stored in the bitmap. The final action of writing raster bits into the bitmap is implemented by an efficient and versatile firmware function called RasterOp. This function copies and modifies bit patterns from an arbitrary rectangular area in picture memory to another rectangular area. RasterOp is described in more detail in (13) and (5). For areas, the run-length encoding can be displayed directly using RasterOp to display each run as a one line high area. To draw curves we use the paradigm of painting with a "brush" moving along their trajectories. The chain encoding provides a digitized representation of the trajectory. A round brush will approximate a line of uniform width equal to the brush diameter. In the simplest implementation of this model, RasterOp can be used to paint an image of the brush at each new raster position along the trajectory. However, because the successive brush images overlap, many of the pixels along the trajectory are written several times. The model can better be implemented by breaking down the brush into a set of horizontal sections and accumulating the rasters filled as the brush moves. Once a scan line is complete, that is, the brush has moved far enough that it no longer touches the scan line, the entire horizontal section can be displayed using RasterOp. This implementation is more efficient than the traveling brush because the affected rasters are only written once. Furthermore, the form of RasterOp used to display lines has been reduced to the same case used for areas. This uniformity makes applying colors or halftones to objects straightforward. Refresh In an interactive graphics system, the displayed image is constantly changing as the user adds, transforms, and deletes objects. In a system which contains only lines and curves, it is possible to write all new objects as they are created or repositioned, and to simply leave the small areas that are erased out of overlapping objects (from a transform or delete operation) unrefreshed. The user can then replot the entire picture when the image degrades too much. For systems which include filled areas, this approach is inadequate because the amount of the picture obliterated by erase operations overlapping other objects is too extensive. It is therefore necessary to find some way to quickly and accurately refresh subareas of the entire picture. We will call this process incremental refresh. The screen update process should be as fast as possible, yet must leave the screen in the correct state as to the shape of the objects and their overlap order. It follows that the design considerations for an incremental refresh algorithm are: The definition of an area to refresh should contain a minimum number of rasters that need to be regenerated. Objects within the refresh area must be replotted quickly and with no rippling effects. If the incremental refresh is going to run as a background process, the time necessary to refresh an area must be bounded. In general, the affected area has such a complex boundary that determining exactly which rasters fell inside the outline would be too time consuming a process. It is therefore necessary to use some approximation to the area. The simplest approximation is a rectangle. While for certain operations, such as erasing a diagonal line, the minimum rectangle that describes the area affected covers far more of the picture than actually need be redisplayed, rectangles are much easier to manipulate than other shapes such as trapezoids. Once a refresh area has been defined, the objects which are affected must be found. This is the same process as hit testing, except that each object must be compared to the boundaries of the refresh area instead of a small area around a point. If the display system contains an efficient clipper with variable clipping boundries, the update problem can be solved simply by setting the clipping region to the boundries of the affected area and refreshing the entire screen. In the type of system we are describing here, since the rasters are generated from the display file, the partitioning of the display file into bounded chunks provides a method for fast culling for this type of refresh. If the clipper is not used, the problem of rippling effects can arise because the object or chunk definition will in general generate rasters outside of the refresh area, which can affect objects not currently in the list of objects to be refreshed. For example, in figure 6a, object A is shown as overlapping object B. Part of object B needs to be refreshed (figure 6b). In refreshing object B, care must be taken that the correct overlap order is maintained between A and B. If overlap order is determined simply by back to front refresh of the display, just replotting all of B will result in B appearing to be on top of A (figure 6c) unless figure A is also refreshed. = Figure 6: Rippling Effects of Refresh. = It is common for the user of an interactive graphics system to operate on several objects in a picture at one time leaving several areas needing to be refreshed. These objects may or may not overlap. One approach is to simply accumulate a maximum area as each object is operated on. However, if the objects are disjoint, it can give a very bad estimate of the affected area. Another approach is to treat each object independently. However, if the objects overlap, intersecting refresh areas will be redisplayed several times, which, while leaving the display in a correct state, is distracting to the user as well as time consuming. A third approach is to accumulate a refresh area for each object, and then process these areas to eliminate overlap cases. In a system where RasterOp is the limiting factor for display operations, the third approach is far superior to the other two. Area Scan Conversion Polygon scan conversion is described in detail in chapter 16 of (5), and much of the terminology in this section will be taken from that source. Here we would like to outline an approach that has been shown to be adequate for line and curve bounded areas, including concave areas, areas with holes, and areas with twists. The problem is to display solid areas which have overlap order but no depth. That is, areas do not intersect in the Z direction. The areas are bounded by combinations of spline curves and lines, and are not strictly convex. The pictures displayed are of moderate complexity, probably around 20-25 areas. Overlap order is resolved using the painters algorithm. That is, objects are displayed in overlap order, back to front such that front objects simply "paint over" objects that are behind them. The outline of each area is chain encoded. Each chunk is constrained to be monotonic. Scan conversion occures at display resolution, using the chain encoding as the edge definition. Each area is taken separately, and the outline is broken into a list of "edges" which are monotonic in the scan direction. This is determined by examining the chunks of chain encoding. These edges can be sorted on Y, then the intersections for each scan line are determined by running along the encoding. The X values are sorted, and taken pairwise to define the filled interior of the object. This is essentially the Y-X algorithm described for polygons in (5). Care must be taken when defining the edges that endpoints for edges are properly defined. The chain encoding is one continuous stream with one definition for each raster point. However, it is essential for the Y-X algorithm to have an even number of edges for each scan line. Therefore, the endpoint of edges that fall at maxima/minima of the object must be doubled. Furthermore, points which are not at a maximum/minimum in the scan direction must not be defined on two edges. This can be achieved by making edge boundries only at maxima/minima in the scan direction (figure 7). === Figure 7: Edge Definition. === It is important to note that there may be horizontal sections inside an edge (figure 8). Therefore, it is necessary to return a range of X values for each edge intersection at each scan line. The left or right value is taken depending on whether the edge is a left or right edge. This determination takes place after the X values are sorted. The main advantages to this approach are speed and consistency. All curves are converted to rasters using some standard algorithm. Once this is done, any analysis of the curve, such as for monotonicity or for intersections, is done using the chain encoding. Furthermore, if the area is outlined, the outline and the edge of the filled interior are guaranteed to match since they come from the same digitizing algorithm. === Figure 8: Horizontal Sections Inside Edge. === Hit Testing By the term "hit testing" we mean given some target point, usually the display coordinates of a cursor, which is the selected object? In hardware augmented systems, hit testing is done by the clipping and display system. The clipping boundry is set to a small window around the hit point, and the entire picture is refreshed. An object falling inside the window is returned as a possible hit. In a raster system, redisplaying the object list is too slow, and redisplaying the rasters provides no structural information. However, the segmented encoding provides both structural information and a way to quickly determine which objects are candidates. The frame information on the encoding chunks provides a method for quickly culling out those segments which do not fall near the target point. The remaining chunks can be decoded using the same routines which display the encoding, except that the resulting points are compared to the target point instead of being displayed. Once the objects which lie near the target point are identified, some algorithm must be applied to select one of the objects. Some considerations are: absolute distance from the point, overlap order of the objects, prefered objects, etc. Conclusions Using these design principles, it is possible to make an interactive system which uses a raster display for design, yet has a geometric data structure which can be used to generate quality output on a high resolution raster device such as a film-recorder, a photo-typesetter, or a laser printer. Two systems using these principles have been designed and implemented on the Alto personal computer. One provides only lines and curves, the other also === Figure 9: (a) Pine Tree by Teri Pettit Figure 9: (b) Gray Head by Dave Bickford. === provides the capabilty for filled areas. Typical imagery is shown in figure 9 and in the illustrations in this paper. Acknowledgments Several people have contributed significantly to the design and implementation of the systems this paper is based on. The authors would like to thank Rick Tiberi, Geoff Brown, and John Dawson for their help and ideas. Bibliography [1] Baudelaire, P., Flegal, R.M., and Sproull, R.F., Spline curve techniques, Xerox Palo Alto Research Center Internal Publication, (December 1977). [2] Erdahl, A.C., Displaying computer-generated half-tone pictures in real time, University of Utah Computer Science Technical Report, Salt Lake City, (1969), 4-14. [3] Freeman, H., Computer processing of line-drawing images, ACM Computer Surveys, Vol. 6, No. 1, (March 1974), 57-97. [4] Laws, B., and Newman, W.M., A gray-scale graphics processor using run-length coding, Proceedings of the IEEE Conference on Computer Graphics, Pattern Recognition, and Data Structures, (May 1975). [5] Newman, W.H. and Sproull, R.F., Principles of Interactive Computer Graphics, 2nd edition, McGraw Hill, New York, 1979. [6] Pitteway, M.L.V., A simple data compression technique for graphics displays or incremental plotters, Symposium on Computer Processing in Communications, (April 1969). [7] Shoup, R., Color table animation, Computer Graphics, Vol. 13, No. 2, (August 1979), 8-13. [8] Shoup, R., Some experiments in television graphics and animation using a digital image memory, Presented at the 13th Television Conference of the SMPTE and published in Digital Video, Vol. II, SMPTE, Scarsdale, New York, 1979, 88-98. [9] Smith, A.R., Paint, Technical memo #7, Computer Graphics Lab, NYIT, Old Westbury, New York, 11668, (July 1978). [10] Smith, A.R., Tintfill, Computer Graphics, Vol. 13, No. 2, (August 1979), 276-283. [11] Sproull, R.F. and Newman, W.M., The design of gray-scale graphics software, Proceedings of the IEEE Conference on Computer Graphics, Pattern Recognition and Data Structures, (May 1975). [12] Sproull, R.F., Raster graphics for interactive programming environments, Computer Graphics, Vol. 13, No. 2, (August 1979), 83-93. [13] Thacker, C.P., McCreight, E.M., Lampson, B.W., Sproull, R.F., Boggs, D.R., Alto: a personal computer, Xerox Palo Alto Research Center Internal Publication, (July 1979). RasterGraphicsTechniques.tioga [Artwork node; type 'ArtworkInterpress on' to command tool] [Artwork node; type 'ArtworkInterpress on' to command tool] [Artwork node; type 'ArtworkInterpress on' to command tool] [Artwork node; type 'ArtworkInterpress on' to command tool] [Artwork node; type 'ArtworkInterpress on' to command tool] [Artwork node; type 'ArtworkInterpress on' to command tool] [Artwork node; type 'ArtworkInterpress on' to command tool] [Artwork node; type 'ArtworkInterpress on' to command tool] [Artwork node; type 'ArtworkInterpress on' to command tool] [Artwork node; type 'ArtworkInterpress on' to command tool] "docgraphicslucida" style202 firstFolioJInsert lastEditedIunleadedMarkcenterVersoHeadersKcenterRectoHeaderKcenterVersoFooter""KcenterRectoFooterK footSeparatorArtworkRule ArtworkClassRuleItitle**I pagebreakL**Iauthors&iEIabstractb OInotefootCChead Ibodyj  RR  YIitemCharSetslSbSR dR &&RRRR RR^KtopRuleRuleIcentertopG83.60835 mm topLeading 83.60835 mm topIndent 1.411111 mm bottomLeading Bounds70.0 mm xmin 0.0 mm ymin 82.55 mm xmax 80.78612 mm ymax  Interpress InterpressInterpress/Xerox/3.0 fjkjWB Ƞ%{rjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^3Z^9Picture DefinitionkrjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^)O^, Interactive krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^(C^kDisplaykrjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr33Wy$[(geometric representation)krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr3)Jy (encoded representation)krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr3(@y(raster representation)krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^)M^; Display FilekrjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^>e^# Output DevicekrjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr3>byG(high resolution)krjXeroxResearch RGBLinear2=D=2krjXeroxResearch RGBLinearPG4PkrjXeroxResearch RGBLinear) F * )D) F krjXeroxResearch RGBLinear)GEkrjXeroxResearch RGBLinear5f_Gf5krjXeroxResearch RGBLinear\SG\krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^?M^d85InputkrjXeroxResearch RGBLinearGJG[) ܣ/N  ?OA =a#:   )7J- =:)&u @ߍ.!o h! ?E] C]N  ^ l H@ GJkrjXeroxResearch RGBLinear) R * )P) R krjXeroxResearch RGBLinear)S)))QkrjXeroxResearch RGBLinear>\>>>^krjXeroxResearch RGBLinear> ^ ? >_> ^ krjXeroxResearch RGBLinear? R @ ?S? R krjXeroxResearch RGBLinear6 K J 4J6 K krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr3?Jy_(pointing device)krjXeroxResearch RGBLinear7JvJtJ5JkrjXeroxResearch RGBLinear?R?Q?P?Okkgn==Icaptiontop##KtopRuleRuleRR RRKtopRuleRuleTtopG76.90556 mm topLeading 76.90556 mm topIndent 1.411111 mm bottomLeading :0.0 mm xmin 0.0 mm ymin 89.95834 mm xmax 74.08334 mm ymax  Interpress +Interpress/Xerox/3.0 fjkjWBr%WBrjBQxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrhPOMj.kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr`POMj.kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrXPOMj.kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr0POMj.kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr8POMj.kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrPOMj.kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr@POMj.kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrPOMj.kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrPOMj.kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrKMj0kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrKMj1kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrKMj0kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr KMj4kxjXeroxResearch RGBLinearH5PkxjXeroxResearch RGBLinearHKPkxjXeroxResearch RGBLinearxK5kxjXeroxResearch RGBLinearK5kxjXeroxResearch RGBLinearK5kxjXeroxResearch RGBLinearK5kxjXeroxResearch RGBLinearK5kxjXeroxResearch RGBLinearK5kxjXeroxResearch RGBLinearK5kxjXeroxResearch RGBLinearK5kxjXeroxResearch RGBLinear(K5kxjXeroxResearch RGBLinearK5kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr0|jquadrantkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr0|jquadrantkxjXeroxResearch RGBLinearPv3vs;Ґs|u'9s_YXpLYyX)y Ρ{oT|XΡQT |GukxjXeroxResearch RGBLinear(̎(kxjXeroxResearch RGBLineartkxjXeroxResearch RGBLinear8׏hkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrMj1kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrCMj4kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrp Mj1kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrP Mj2kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrPMj3kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrpMj4kxjXeroxResearch RGBLinear`0kxjXeroxResearch RGBLinear8kxjXeroxResearch RGBLinearkxjXeroxResearch RGBLinear("(kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr`ؠj QuadrantskxjXeroxResearch RGBLinear 莘kxjXeroxResearch RGBLinear kxjXeroxResearch RGBLinearLpkxjXeroxResearch RGBLinearK4җkkkg==Utop--KtopRuleRuleRK 144 pt keepRuleRuletopTG81.66804 mm topLeading 81.66804 mm topIndent 1.411111 mm bottomLeading :0.0 mm xmin 0.0 mm ymin 78.84582 mm xmax 78.84582 mm ymax  Interpress JInterpress/Xerox/3.0 fjkjWB}%{rjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^/M^31krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^GR@^38krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^BY ^31krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^9U@^32krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^. ?^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^- ?^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^, ?^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^?^r.krjXeroxResearch RGBLinear;@ krjXeroxResearch RGBLinear/ krjXeroxResearch RGBLinear2 krjXeroxResearch RGBLinear(WakrjXeroxResearch RGBLinear#\-krjXeroxResearch RGBLinearp )ޥ W?%[ krjXeroxResearch RGBLinearp )%[ W?ޥ krjXeroxResearch RGBLinear9[8 \ [ 9[krjXeroxResearch RGBLinear9`g80;9`krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr3(Vyf1OctantskrjXeroxResearch RGBLinear4[8_krjXeroxResearch RGBLinear4[8krjXeroxResearch RGBLinear*J 0krjXeroxResearch RGBLinearL;KQV# 3krjXeroxResearch RGBLinearBV@HkrjXeroxResearch RGBLineare:F)krjXeroxResearch RGBLinear<TBkrjXeroxResearch RGBLinearc)K#3krjXeroxResearch RGBLinear3Kֶ# s)krjXeroxResearch RGBLinear6 M6 SkrjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^,_@^31krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^* b^32krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^'@b^33krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^$_@^34krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^$\^35krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^,[^38krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^*@Y^37krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^'@Y^36krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^+G^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^*G^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^)G^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^@G^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^AG^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^7G^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^BG^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^6G^r.krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^5G^r.krjXeroxResearch RGBLinear,? krjXeroxResearch RGBLinear.? krjXeroxResearch RGBLinear0? krjXeroxResearch RGBLinear2? krjXeroxResearch RGBLinear<? krjXeroxResearch RGBLinear>? krjXeroxResearch RGBLinear:? krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr36Cyr bitskrjXeroxResearch RGBLinear+?2?9?@?krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^1 >^31krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrK^7>^-hrkrjXeroxResearch RGBLinear+<@2<@9<@@<@krjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr30`;@yLGoctantkrjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr3:;`ylengthkrjXeroxResearch RGBLinear %J`ӁF;Ի+ J@. E;5L5EE4'LUnXC;UgWrIgmWMyE>uoVj;:+r3@]<U{k[C%v4S1"K3O, Wj,C>'LUnXC;UgWrIgmWMyE>kxjXeroxResearch RGBLinearkxjXeroxResearch RGBLinearx08 ŗ)DQOf83x+"U5qrK1H)&A>;SԫAT'?u8pWV0%"c5aOF5{:sx0x08 ŗ)DQOf83x+"U5qrK1H)&A>;SԫAT'?u8pWV0%"c5aOF5{:sx0kxjXeroxResearch RGBLinear808kxjXeroxResearch RGBLinear8 kxjXeroxResearch RGBLinearXkxjXeroxResearch RGBLinearkxjXeroxResearch RGBLinear-:52kxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr$B3o5SkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrH fo5SXkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrx po5DXkxjXeroxResearch RGBLinearkxjXeroxResearch RGBLinearm<܎kxjXeroxResearch RGBLinearmLUmkxjXeroxResearch RGBLinearU=NUkxjXeroxResearch RGBLinearL=%=LkxjXeroxResearch RGBLinear% D%kxjXeroxResearch RGBLinear8XkxjXeroxResearch RGBLinearXkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrjo5HkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrx蠢o5WkxjXeroxResearch RGBLinear0kxjXeroxResearch RGBLinear񗘠kxjXeroxResearch RGBLinear71kxjXeroxResearch RGBLinearhGS%NSyu9i4NMhYI-KQOY](_(*zghkkkg Interpress:0.0 mm xmin 0.0 mm ymin 122.0611 mm xmax 94.54444 mm ymax G97.36666 mm topLeading 97.36666 mm topIndent 1.411111 mm bottomLeading ==Utop..topRuleRuleVqqV$$VGG R RR R VllVWWVzzRRRKRuleRuletopTtop $Interpress/Xerox/3.0 fjkjWB0 WBrjBQxjwwwXeroxResearch RGBLinearsBT4J7S;"SIe {EiD[@3%v0 ?"S75"?+wczXdva=1usBT4J7S;"SIe {EiD[@3%v0 ?"S75"?+wczXdva=1ukxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr0 d_jo5BkxjwwwXeroxResearch RGBLinear&ʗܗ4җ7S;k'/I xEiDvX3%0 Q"S7s"?+czX)sm81p&&ʗܗ4җ7S;k'/I xEiDvX3%0 Q"S7s"?+czX)sm81p&kxjXeroxResearch RGBLineargB.G!=QF"'}*桒@4UEshUEޡnUqʍԡCR71nawveZ%SӡvByB7q|yA'7ɡsyy-Lw@[ޡzAoX$o7@dwJMQ4v@kr+kkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrx렢o5AkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr0 젢_jo5BkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrȠija) Objects A and BkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr@+jb) Rectangular region erasedkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrjc) Replotting all of B giveskxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrgBjincorrect overlapkxjXeroxResearch RGBLineargr.G9QF"'}j{-@4DEsh%EnUN="䡒CR7!ջaСwveUSN㡒vBy|7q|y{7١s y-w@[zAcX$8o7@OAMQeE*vkr+kkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrxo5AkxjXeroxResearch RGBLineargZ.G-uQF%"'}R-n@4zEshz}EfnUq!\CR7b)SHwve=Sk[vBy_7q|y^_7Qsy-nw@[fzAX$R-o}7@JMQN*vkr+kkxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrrxso5AkxjwwwXeroxResearch RGBLinear̗4—7aS;e SIݡ W(EiD@(3%0 -"S7" ?+czX U=1u̗4—7aS;e SIݡ W(EiD@(3%0 -"S7" ?+czX U=1ukxjXeroxResearch RGBLinearXerox PressFontsTimesRoman-mrr0 ܠ_jo5BkxjXeroxResearch RGBLinearz+z׎kkkg Interpress:0.0 mm xmin 0.0 mm ymin 85.37222 mm xmax 141.1111 mm ymax G143.9333 mm topLeading 143.9333 mm topIndent 1.411111 mm bottomLeading ==Utop&&KtopRuleRuleR RRRRRK 144 pt keeptopRuleRuleTtop Interpress/Xerox/3.0 fjkjWB񠢠 WBrjBQxjXeroxResearch RGBLinear9X=Zd9+!8PCT/?>U\YSR}IS'pMkxjXeroxResearch RGBLinearM9P=I9+!P-/?>r\YR[IS'{pMkxjXeroxResearch RGBLinearMOGBD2u&z'.uMO/_McD@L#U8AEdu8Qa'h}_-&Wi/(QaXYBGykMOd){3@c=/qak,wlhashiJAX%3+@5>vY_44);: X3+'.mdXi)h7K7%A8($2hGhaG|GMOGkxjXeroxResearch RGBLinearaGrM2u.ua/sD|W@ayoEkO8S>'<-_-&ћ[/(S>BiYZykaA+d)s`39]/,wl1iUX%>3p4N,sQg;:e~; X'gu Gi)8a7A828aGGF.GaGkxjXeroxResearch RGBLinear뗘kxjXeroxResearch RGBLinearkxjXeroxResearch RGBLinearkxjXeroxResearch RGBLinearvfkxjXeroxResearch RGBLinearkxjXeroxResearch RGBLinearbRkxjXeroxResearch RGBLinearp|kxjXeroxResearch RGBLinearyn[kxjXeroxResearch RGBLinearQhmCZQkxjXeroxResearch RGBLinearP@kxjXeroxResearch RGBLinear'kxjXeroxResearch RGBLinear7'kxjXeroxResearch RGBLinear0o kxjXeroxResearch RGBLinearkxjXeroxResearch RGBLinearkkkg Interpress:0.0 mm xmin 0.0 mm ymin 89.25277 mm xmax 89.95834 mm ymax G92.78056 mm topLeading 92.78056 mm topIndent 1.411111 mm bottomLeading ==UtopKtopRuleRuleRRKbottomRuleRuleTbottom Interpress/Xerox/3.0 fjkjWBSwWBrjBQxjXeroxResearch RGBLinear0Џ@0kxjXeroxResearch RGBLinear@ЏP@kxjXeroxResearch RGBLinearЙЏkxjXeroxResearch RGBLinearЏ kxjXeroxResearch RGBLinearkxjXeroxResearch RGBLinearkxjXeroxResearch RGBLinear Џ0 kxjXeroxResearch RGBLinearЎkxjXeroxResearch RGBLinearЎkxjXeroxResearch RGBLinearxkxjXeroxResearch RGBLinearxkxjXeroxResearch RGBLinearxkxjXeroxResearch RGBLinearxkxjXeroxResearch RGBLinearxkxjXeroxResearch RGBLinearxkxjXeroxResearch RGBLinearxkxjXeroxResearch RGBLinearxkxjXeroxResearch RGBLinear xkxjXeroxResearch RGBLinearxkxjXeroxResearch RGBLinear0xkxjXeroxResearch RGBLinear@xkxjXeroxResearch RGBLinearPxkxjXeroxResearch RGBLinear`kxjXeroxResearch RGBLinearЙ`kxjXeroxResearch RGBLinear`kxjXeroxResearch RGBLinear`kxjXeroxResearch RGBLinear`kxjXeroxResearch RGBLinear`kxjXeroxResearch RGBLinear`kkkg Interpress:0.0 mm xmin 0.0 mm ymin 81.49166 mm xmax 44.80277 mm ymax C47.625 mm topLeading 47.625 mm topIndent 1.411111 mm bottomLeading ==Ubottom**KbottomRuleRule  RRQ I widowsaverK 144 pt keepRuleRuletopTG80.99777 mm topLeading 80.99777 mm topIndent 1.411111 mm bottomLeading :0.0 mm xmin 0.0 mm ymin 61.80666 mm xmax 78.17555 mm ymax  Interpress ӍInterpress/Xerox/3.0 fjkjWBlT/o%{rjXeroxResearch RGBLinear5\ _se#3^3u)# 1]#T $P 1\H{N&2[ C,, WY3Z@ Pe3  f3 1Yࡒ <E0`X( _k; 0Wj9=  [~ 1V#C#; 2SࡒhY gRM1 PI(  .Q@Fk,'eI- Q` s+Q@# I?# +P  C!: 3,O`Y1pI i+NA-[ nS:*N`qM Q+*MtG0 4 5 , L?"Pd#-L@  q ę+KЅ)-&8 *J_$ 6=kB  o.,I2r/ .H11.zMu.Drk%p  , C owB[ q)D bWzAHW(D : 9 (AM &l)`@! `%*?`` 67'w(>@L)} )`=@K F; * <#H(;`&% +2&;:?%@<_g#<``+Sy.#:`Խ"$9U??u)#8t^1"`8ࡒ58 qb(!`8 %[, +k[!7 ``@ ] #6@$;) ^?$!5U!#!@4BNeS %e"@4 g#@3ࡒ# "nG$`4@w8C 2> #2@/ 'q)+-#@1"(2'/$0ࡒLy &@0E3\8 .&`.  $-=m" .%-)+ @0 sO.0ࡒ,a,_ .pG`ߊG/6# S  `/@Uqyv81`.#z .4 -@kc6}} ,0f|1D,tJ#<G7B +@i!e _G` @*`v*`s8f@7-c* ( SS )@LA!'>) 'ࡒ]#8@a7#a  '@+ ?5 k' /  M/ ( `'`& &@%%h% &%!d)! %iL/I:!&1,Sx $8AN)!"ࡒ  t X "#j$@% Iq kRB%&@J7qAC &@%`#3.u*'$`u+  !o&#`;!:kg%" JuvT&! ,:,D=/'`!3&$. +)`"M3WS3*@$`%5*" 7&%+!f,@ 5k% -!ࡒ N' -@$g.j.$`  hd 5&0$` o]KgK03`"u Y|ce4`#(5S/55$`."# n?6$ࡒƪ?18#ࡒe:s'9 "` d;>CJx: ࡒ;" o 5<#::<$9y ~9 =$s"5>#Sn  7?@"`a>'i@# U Yi @$ O zA &[c B % 2:?D@#@*xU)E"ࡒW Q+F`$~ U.bgE& {iT&I%@lkA87K % VmK&J '@H LK ' u+"!L'T4y ŠN'g؅ EkeP`(( u,N)ࡒ$  ,-K*CvUI`+ 6  3FI-` q8fJ/@5  rA0M`/"y,; N`0ࡒ'0 #`"M2` K1EK 2@qH2&xL 3"Ui  | IL4ࡒ.K5ࡒ21L;CJ`5mѧ>_yI4b~v m}H6 =  H7;,3>T G6e,P ^]LF`5 5=2c4D`5@n`tC6 ݡ,] C9| pE D8/ZQM  E8X3 *F9la" a+" E:L 4= G ;D@;`R ; sa> E;3fG`;L/ bCH;b܂33H =` x wG =; H>%"o H@ rj  ?pG`A qt:O!jE@`xDW Z$E A ) N  ;E@BL  ;H 6yD`CBB ` B@0N  = AB֋Ҭ#) @C9yχe9"?`B>)J)?Aj 5 >@)I I>B`~~i«F> D W  {3rE NF @`E2_:?G@ScX >G@/w P<?Gࡒo $ b A@F> Z f3 7iC FࡒH8   BHࡒ T5 = @`I=ҭN @K@j8>L .pf-=K b =`M'_r ->M }- 3?N ? >`O?UCU 77<`OࡒMg_M ; Oࡒu"; E;Q @ f(Y =R   =SࡒM. 2; .;@T`% ~;f 9S# o # gy9W@5*G 8[? M+B 9[ࡒwdaf5 9@\1 8] =6\6`@6 5\5\ _se#3^3u)# 1]#T $P 1\H{N&2[ C,, WY3Z@ Pe3  f3 1Yࡒ <E0`X( _k; 0Wj9=  [~ 1V#C#; 2SࡒhY gRM1 PI(  .Q@Fk,'eI- Q` s+Q@# I?# +P  C!: 3,O`Y1pI i+NA-[ nS:*N`qM Q+*MtG0 4 5 , L?"Pd#-L@  q ę+KЅ)-&8 *J_$ 6=kB  o.,I2r/ .H11.zMu.Drk%p  , C owB[ q)D bWzAHW(D : 9 (AM &l)`@! `%*?`` 67'w(>@L)} )`=@K F; * <#H(;`&% +2&;:?%@<_g#<``+Sy.#:`Խ"$9U??u)#8t^1"`8ࡒ58 qb(!`8 %[, +k[!7 ``@ ] #6@$;) ^?$!5U!#!@4BNeS %e"@4 g#@3ࡒ# "nG$`4@w8C 2> #2@/ 'q)+-#@1"(2'/$0ࡒLy &@0E3\8 .&`.  $-=m" .%-)+ @0 sO.0ࡒ,a,_ .pG`ߊG/6# S  `/@Uqyv81`.#z .4 -@kc6}} ,0f|1D,tJ#<G7B +@i!e _G` @*`v*`s8f@7-c* ( SS )@LA!'>) 'ࡒ]#8@a7#a  '@+ ?5 k' /  M/ ( `'`& &@%%h% &%!d)! %iL/I:!&1,Sx $8AN)!"ࡒ  t X "#j$@% Iq kRB%&@J7qAC &@%`#3.u*'$`u+  !o&#`;!:kg%" JuvT&! ,:,D=/'`!3&$. +)`"M3WS3*@$`%5*" 7&%+!f,@ 5k% -!ࡒ N' -@$g.j.$`  hd 5&0$` o]KgK03`"u Y|ce4`#(5S/55$`."# n?6$ࡒƪ?18#ࡒe:s'9 "` d;>CJx: ࡒ;" o 5<#::<$9y ~9 =$s"5>#Sn  7?@"`a>'i@# U Yi @$ O zA &[c B % 2:?D@#@*xU)E"ࡒW Q+F`$~ U.bgE& {iT&I%@lkA87K % VmK&J '@H LK ' u+"!L'T4y ŠN'g؅ EkeP`(( u,N)ࡒ$  ,-K*CvUI`+ 6  3FI-` q8fJ/@5  rA0M`/"y,; N`0ࡒ'0 #`"M2` K1EK 2@qH2&xL 3"Ui  | IL4ࡒ.K5ࡒ21L;CJ`5mѧ>_yI4b~v m}H6 =  H7;,3>T G6e,P ^]LF`5 5=2c4D`5@n`tC6 ݡ,] C9| pE D8/ZQM  E8X3 *F9la" a+" E:L 4= G ;D@;`R ; sa> E;3fG`;L/ bCH;b܂33H =` x wG =; H>%"o H@ rj  ?pG`A qt:O!jE@`xDW Z$E A ) N  ;E@BL  ;H 6yD`CBB ` B@0N  = AB֋Ҭ#) @C9yχe9"?`B>)J)?Aj 5 >@)I I>B`~~i«F> D W  {3rE NF @`E2_:?G@ScX >G@/w P<?Gࡒo $ b A@F> Z f3 7iC FࡒH8   BHࡒ T5 = @`I=ҭN @K@j8>L .pf-=K b =`M'_r ->M }- 3?N ? >`O?UCU 77<`OࡒMg_M ; Oࡒu"; E;Q @ f(Y =R   =SࡒM. 2; .;@T`% ~;f 9S# o # gy9W@5*G 8[? M+B 9[ࡒwdaf5 9@\1 8] =6\6`@6 5\krjoXeroxResearch RGBLinear6^ 6^61@6^ 6^ 6^61@6^ krjtXeroxResearch RGBLinear6`a%@"C 5`` :@? B@? ~ 5`^G+|yy 4`_ke_3`__sW- 2`^;|t/- 3`] /   4`\=L s? 2`[tI0%szq?0`[k/!0`ZTo 3`XX$1`X $lF 3`W  wp6`V4 oD!e49`W׿`;`Zl_i;`[\$:`[ͩ _t 8`[ij &  9`]9 K ;9`^F5 B7`^w0^w _@7`_w``xa`7`avaupa6`a6`a%@"C 5`` :@? B@? ~ 5`^G+|yy 4`_ke_3`__sW- 2`^;|t/- 3`] /   4`\=L s? 2`[tI0%szq?0`[k/!0`ZTo 3`XX$1`X $lF 3`W  wp6`V4 oD!e49`W׿`;`Zl_i;`[\$:`[ͩ _t 8`[ij &  9`]9 K ;9`^F5 B7`^w0^w _@7`_w``xa`7`avaupa6`akrjtXeroxResearch RGBLinear5`V, H 3.Ww w-VW -Wࡒm"#z+X   @s+@W`X;  kz8+Vo  q Hk ,@U  ` *`V)pR A(V4 )'V`Kzj  8 ' U`:6(T !)T+P:+(T(S.+lc2 (Sa([ &Uݿ (R (*i(o* R  +}X^ g[_ +@So y* Ax+R E Kg_* QV e(t?= *P4}G ,@P K-Q@$3^3`.@R`P{hb-`Q E ֹ-@P@53….P»Y[/P . O1o-9-Mࡒ['/. M W  /M Y 922 /L`T   GU/KfV p 401 K1[ 1D 2 L@ L 2@Kc;Sc O  B52JR 3 3I`N!0=?`O@^*'C @`P` ƌ3U{.R AQ@)  __@Q  ?@QࡒFLmWi @@R`W! 3HAS '3 B T@V Q*" AU  DM   @T *#x)&?@UHNc >`V@mf{ Bk=U@AB AW=T߻ReM =@RyY sw ;S`{{ W: S 7,ea ;`T  XoQ )<`U@fީ 9 <`Vn /s:V- , ( V 9V  QG?3"8W M8@Wࡒs3t 7X ؍ PWD6W`4  6`V6 #s6`U kS7 5`W  1[Q, 4`WH @ V 4 V   -4`TࡒT4`TH, T3VeD .#C2@V m"1 Uࡒ3+ .<I 0@Uࡒ& ), D'/VE>3.Ww w-VW -Wࡒm"#z+X   @s+@W`X;  kz8+Vo  q Hk ,@U  ` *`V)pR A(V4 )'V`Kzj  8 ' U`:6(T !)T+P:+(T(S.+lc2 (Sa([ &Uݿ (R (*i(o* R  +}X^ g[_ +@So y* Ax+R E Kg_* QV e(t?= *P4}G ,@P K-Q@$3^3`.@R`P{hb-`Q E ֹ-@P@53….P»Y[/P . O1o-9-Mࡒ['/. M W  /M Y 922 /L`T   GU/KfV p 401 K1[ 1D 2 L@ L 2@Kc;Sc O  B52JR 3 3I`N!0=?`O@^*'C @`P` ƌ3U{.R AQ@)  __@Q  ?@QࡒFLmWi @@R`W! 3HAS '3 B T@V Q*" AU  DM   @T *#x)&?@UHNc >`V@mf{ Bk=U@AB AW=T߻ReM =@RyY sw ;S`{{ W: S 7,ea ;`T  XoQ )<`U@fީ 9 <`Vn /s:V- , ( V 9V  QG?3"8W M8@Wࡒs3t 7X ؍ PWD6W`4  6`V6 #s6`U kS7 5`W  1[Q, 4`WH @ V 4 V   -4`TࡒTkrjtXeroxResearch RGBLinear3`?2`A 1`A0@0A/C@-B,B+D *C*C B*@)A`(`B`'`B&B'@@%A$B #A$@"A AA@ @` @ !? > =!<"=">#<$@= % =%<% ;& ;'@< &:' :(`:@):*@9`*8`, 8-9. 8.70 70818`173@7`4868`7787989`:;9<9@:<@;`> ; ?;@;?A? @>@`?@?`>`@ =?A=B =!<"=">#<$@= % =%<% ;& ;'@< &:' :(`:@):*@9`*8`, 8-9. 8.70 70818`173@7`4868`7787989`:;9<9@:<@;`> ; ?;@;?A? @>@`?@?`>`@ =?A=B`E`=F=F??G@@H ?H>A I@`J? J>@K`=@K`E`=F=F??G@@H ?H>A I@`J? J>@K`=@K@$ >&@%`A@$ C $@D%@C&`B '@D&F&F (G'I@'J@( I)@L(M) M*O@*`P*P+Q@,@Q-O.N`-`N .@M/@L`.K-`K@.J/I.I-`H@. H`/ H0`G`0H1`2F2E1D0D1B1@B0B0A 1 @@0 ?@/`>0@=0``9=: <`:; : :8`:`7@989@887@7 7 65`7@4`73 638 38 2@9@1@81 605`/6.`7@-6 - 7 +7*7 5`)6(8'`7&8@&9% 7%5&5& 3&2%`2`%@1`% 0 $`1#`0# 2"`3!2!1 3 @5 4 `5`65@6@76@5 6  5 @4`3@2@@10`01@ //``//. - -- ,@+!,! *!)#*`$*$)#)")`!@* * ) )`)' '` &`%$$#``"@@" " ` !!@# "@"##`$$$`%`$@&% &&'% (`$(%(&)%*% +%+@',@&,`% ,#`.@". !/ ! 0`"0#@1" 2!`34"4!6!`6"6`#7#8#8%`7&8&9`% 9$@;$ ;% <@#=@# >@$ >&@%`A@$ C $@D%@C&`B '@D&F&F (G'I@'J@( I)@L(M) M*O@*`P*P+Q@,@Q-O.N`-`N .@M/@L`.K-`K@.J/I.I-`H@. H`/ H0`G`0H1`2F2E1D0D1B1@B0B0A 1 @@0 ?@/`>0@=0``9=: <`:; : :8`:`7@989@887@7 7 65`7@4`73 638 krjXeroxResearch RGBLinear/P 0O0 P0R1R`1Q2S@2T 3SkrjXeroxResearch RGBLinear3@N2M4 M 5M5N6M8`M9`M9 N:O krjXeroxResearch RGBLinear0`@0 >/ ?-@,@`,@?*?`*? )`>@*@=@+ <(`=@(<(;krjXeroxResearch RGBLinear= ?< >;`@@:@9@?8@krjXeroxResearch RGBLinear.<-=+<,@;`krjXeroxResearch RGBLinear3`:49`6 9 6:`79 krjXeroxResearch RGBLinear8?`8> 8 =6>`5@?4`=krjXeroxResearch RGBLinear1-0//@0./ ,0`+1*`0*/(`0 '0&/krjXeroxResearch RGBLinear&,%`,%*`&`)`'krjXeroxResearch RGBLinear!' &%@!`$"`%krjXeroxResearch RGBLinear+`)*(+ ','`-'-%-$/`$/%1$2$3&krjXeroxResearch RGBLinear2/11`13224143547 372`8 49@6`:6;`5 <7=@8@>87 krjXeroxResearch RGBLinear2`31@0 krjXeroxResearch RGBLinear91`;2;1 ;@/krjXeroxResearch RGBLinear5( 7'@9'9)`krjXeroxResearch RGBLinear9`)<`(>@)`=*;,= ,=`. <@.krjXeroxResearch RGBLinear?+?@- @.A.`A,B-D/E@.`E ,`F.krjXeroxResearch RGBLinear4 ,4`.@5.krjXeroxResearch RGBLinear@(A'`B( C )krjtXeroxResearch RGBLinearD?D@@B@B?krjtXeroxResearch RGBLinearI@2J@3 K2@K@1krjXeroxResearch RGBLinear- ::`99  /@..@- - ::`99  /@..@- krjXeroxResearch RGBLinear.9kkgtop==Utop&&TG89.95832 mm topLeading 89.95832 mm topIndent 1.411111 mm bottomLeading 90.0 mm xmin 0.0 mm ymin 76.90554 mm xmax 87.1361 mm ymax  Interpress Interpress/Xerox/3.0 fjkjWBzo%{rjXeroxResearch RGBLinear@X ioI B`m@%2= ;AR'[mh r_ {Ӹix m 0$ r U`~A  Rfs1 JH\0e @XkrjwwwXeroxResearch RGBLinearsx?0NIm  =35P?ࡒh, e34H?k\mHPڶ8" M"p`>D# sx?0krjwwwXeroxResearch RGBLinearV0mh)!b0LHإ[T_#\ GxQhc  J|CXN3 5" '8XbFN r'6G >3[8 B IQ!ԗH *Ph 8+AkH m UgT9K ?z -h,`~ 9kB?SH(7Ho |1?AHZD- @>,Ht 0"]$]T|KsJ\ U@DxCj mgX`0g"C8!HlXWs-#X:Qp8  x km6v xz ) hȡ o<=  o Ʒ ge!W dkrjwwwXeroxResearch RGBLinear5Ph|AOA koXpL  I )D=4,4D?$ p`E "y, qavu( ?Csx+} sxeN  G6E`,O|* 7`8"]  | S1 8~  8hG =7{Fh W  8U[{% zr*N wVAz+ h 6DXx  - sxDs&ACsxxA D sxI5PhkrjXeroxResearch RGBLinearFphn  E@@< FpPg HIءI> 0'LHpn+J XJ57 gpx%2sgPpk8@`pm77#I > ,MU"2(@˴ Y)4  Fphkrj===XeroxResearch RGBLinearl@>ؙe $n>X>t) p`>3} rp Mv UbG;(H+sP C;(/&ZO}0ءl@>ؙe $n>X>t) p`>3} rp Mv UbG;(H+sP C;(/&ZO}0ءkrjXeroxResearch RGBLinear  uI ExQh ,F  A3yOm3X/ ?{c8Q ^g  tƸV D$UG sP#M ) v0&@za C  |m7~{DX*m  *6x9 } tCN/ : X!IA#!@K3S  > P?ࡒmP  ;N _I ?@?0r IXgc  4 UN`d 7w$  3Qh =e`  5PkrjXeroxResearch RGBLinearxdPjGM  "8y 4W ^goX m d\-C" Ra e- J8\h='א)d @'Uࡒrƒ I+5rh W+X cU*E.pU0c  * 10#S@*G2Rȡn3cq ׼ i4 e 3@Pk2{h d l@/ l$D ]kH {+7 jF{\ h \'mHݣ  40P0{o m Mv rpE3EsK5P1(9(58e (= 6x_ R - -5P#T sx) c ? @Ht.Rpx# + BN 7 ߳ F(qW q)VX ].4 g16KD@=; 6/ [  F IdP/N =xdPkrjXeroxResearch RGBLineardXV 0ƀ0(R  h .^) =%V"t0 a *QWB>gZMPS+=iϬMPP`O EE@{ *P}y TO_ 3_79( .x[=;3a@OX?i  0P`A1b 8i dXkrjwwwXeroxResearch RGBLinear1V3# e  ^(^(K8|jec/fx,K8G j  ((K衒)5+ w_qH(Jzi% #K0? ;+Xx 9ŭ#k,6A>}a,` .X> M? GbHH5@1j e`Esd  {#U b=gphl?m1krjwwwXeroxResearch RGBLinearJ85>E 1LH  gLHܸ1 2Mf!W  ,z e;ZZ{-XA \wR9: / YP }-c[0(l@IgR 2}"  2Sxh @; ] Hw:  +w-U( s Zo J8krjwwwXeroxResearch RGBLinearuXԍM3 ppv'   0 q8/c dsSxg7wS & ,MxA&~|OJ3 .|0N "KWOMJ. y E"r= E T(Q?   U AV8 *bUpk_:t}X'  j {a q U Rxڨ D$mn.uXkrjwwwXeroxResearch RGBLinear5<șiL+L o(]_ ,NGQmH= S_ ,k85C !Lk8 ; Un;hev KI<sx7@ _'w;M1 8E  nE|<ȡ i z<ȡί} ylru280~x !thxRp.h,-2_7,#Gl@}+mHX)H%k8H OA1C kf"j!ASHk@'lkmHEn+"4*Eoi E4zbêox(8*o 8p`C+iS4 AK x 2 3 ^~3 idt fhA8?mqG)feQ[R Uq  1 H^Sx gpIء8MT e`2e w>qg[ep_ ye`P` ȭ U bsY~_n]  C2`z,`(ݳ[8Ue- =P{?5!K@^J Uk_ GxRp+4_ k 5.'(krj===XeroxResearch RGBLinear|Rp]w O "8S  ] 'v, xT(M+8Z!&m 9pP!|n ͌ (U0)y \M h`h$O,< Ub<kUࡒa/ 3- % ( yIk B 8 %% PD! ~s$& w2=TkO - mۣ>?HSxeq _9 Ɏy/(# E.9k|Rpkrj===XeroxResearch RGBLinear: RkY ^e2H(6R  SxrH SxS{a x! * 2*#  RT ѾVs <BQh )@QGD*U"AQ C?T M)xQ` I\WJ8 hGK*% 6Y  >@S [ew] XRȡO5 d 0 N2 $I: RRkrjXeroxResearch RGBLinear0Ƹ, d " it <+#"x(D,H>(GN!no  Xue   ?@G@PŰiP `àGku>@P`C0=`˜@0 (Xц 0ŰZ Ui 0Ƹkrj===XeroxResearch RGBLinear0(JX/ Ƹ_ ˘A .ZF Mq#URpE n)QpƸ-I(y һ/qNXQlaC'8 ja2NX"2O BH H E LH˜8hk)H = 93)S<dS 3IOO'@ 'O'%G 3TQ!'˜xf8Kw ]/PP^ 2O , `3HD h Z %#W #c/ ihZ 0()``x)4H0(krj===XeroxResearch RGBLinearaĨT;Pcb˜|% %}Jb L $B#0(O@?/x5< 2"/ x1Xa l a@OXW"> H 0 10 FK0k ȥ0ܝ0Ĩ j'TE aĨkrjwwwXeroxResearch RGBLinear`S q!X>]\;T)4tI 'X0= R )>?%#;h%iڶD}+;hOU>J H8Tx]KeOk WIHСL \KeH0hgۭ CFh  'VOX*[OX U XQa`8փ^ ]x+fu}-kCms(6F S00N+)T  PhOOkrjXeroxResearch RGBLinearUQhN ).,`*CN # O.,˜˛T-hΩٗK ZĨJN GF @YQɭj5ϐ,Qw> 1t  UQhkrjXeroxResearch RGBLinearYƸϘ` 1-ŰGP ' \Yz!Q &,ZQhJ= #73zfZYƸkrj===XeroxResearch RGBLinearBdO Lo?4y B7 >  0| !uzCh  @ ߝN W~w G +f#" ;e J2Q&P` 2Ɵw* L,C @1 aq-HD %S=G1 mP +/b)x}jGUk GB7P@衒o  H†'_ Et $98= 6MExp/g ܑ *Dp A@\) ?l } q"-; P ?)AP 7W haqd2d?;(!?Q m u9x "ȫ =)J58_E} Q;hM4h4 H V Hx\9 X9H   O `NP )!ԣP ].v<8) `01< Pd#LJ _cE hp]8 b?@B` yD5 ?! g3BdkrjwwwXeroxResearch RGBLinear>( Fy$Z :xM u#| i$v;\~ r us?f x~>' w  ~}E/&9X36:7 ?& u k;(s_?i<6i#> x>@9Izk d  8;=S3(:ip/"RF W >(krj===XeroxResearch RGBLinear+*j0b|, 7~mH  z ]+sx:3D.uu  hI@8UzdXM`Mi;"'O`/`CF ;LE*?8m#,Fp֒~D ++&ua,4b +40yQ'897'+Ȭ=R  [([ s 5]v" /krjXeroxResearch RGBLinear0`esy!!( 4 4?0 O SGI@ 9!=  ?)c)c(#2X]w/0`0krjXeroxResearch RGBLinearhql]),P+="H8)5^X( W" `@Cm @  P?ࡒa hkrjXeroxResearch RGBLinearh8k 6 x 3+<-  5 0Hatj0 , >(e i G}f3OZ BHsQS $h8krjXeroxResearch RGBLinearEx 5Y 4 >(#e/ B:‘ G  1%zEx<ȡ xv-O RD AZ@~1S+](?ࡒ %u NqExkrjXeroxResearch RGBLinearf +2>(ۑb ;/ 3>(#;[ ] p`8; 6=СGG?u>2b iN#x>" z"ikm#|>Q : @$  WO|xM. '60 <'5'5P0ub& kX k8_;AePB gp0y= [,fkrjXeroxResearch RGBLinear^V ʣJ9>@!E'H;cQS= ~ b@ rpx)1 ~v 0hemc38Sh.p_G$>P`ࡒ.  Vx6)>fIi/ 0,+`0Ue`*B Y~l#*7* ! 7`b )~ 7U @  =8fW krjXeroxResearch RGBLinearMY om 49m W\+ p . e&r ;u   !-NPb  uGȡϹ# T LXy Nh 0$ H? < N }L ?X a? & MkrjwwwXeroxResearch RGBLinearMUMޏ M`HZb  UK衒 H'Cq`i  xX  5H( p`({k@?xK衒 IL5h (  M`Q4 & ~ Mkrj===XeroxResearch RGBLinearRpe )B$`g &CUpP`G_J 0 2̇ 6}@ĨJlw l  (J3HV-  >eG  Y. ] >@ d?S=8 j, `Rȡ ttrRpkrjXeroxResearch RGBLinearhRXU '>< H9W !  .=8Ĩ;đ=gJR Ƹ 'G\ @0RpU 5`=8RpM  K? !PhRkrjXeroxResearch RGBLinearh& A=GVB ́ h!1 ,R]  FG3r krj===XeroxResearch RGBLinearmHC<28h>!#bHq =" `8@衒MF,U-\P "F]6>ZH -*?4W-" G1Cb^( 3/e`z= t7* %jH  X@#l@X@M= ;nPnTh.h p`*"'yr`J#^ U7`dC9H]u)>(ٙ.m gt=СqFsءb   =( "t'8 AZQ]{PeVouhz{ 0W g5P : HLmHCkrj===XeroxResearch RGBLinearoXomYqh-xax )6E.  (t uE`(TV)kSv @G vD+-uVe6X(N  7rpE`&oXkrjwwwXeroxResearch RGBLinearh o %{Dg Flx-g5 ց% 028< l i(DcA B! hxqc$ E4h y !¡ yW h krjXeroxResearch RGBLinear1<șs Kz)mg;cAS!@5Q'ej0;hǃ?1A  l@j J ?4:`um y  @rp:` X&;5:_t:`8VyC.`wEֱ)z:L  RS}U~ ~"y '\ 9y` @yxX != qhwPT SN Ii d8e_ &7_0~2?e b<g 1<ȡkrjXeroxResearch RGBLinearN\x 5 + j\^SC N 7ISСvO \[  7su 1 nk PA 9|4Q#b *K  O F \ t3<Mh } <^ >0 = u<3 Nx@  : HA@ K.*Us PT(ʐ6 i Z<e ]\z H 3N\krjXeroxResearch RGBLinearPMHI   0Ź  D)  X*&!_ pOpK8   #(   5 % 'r M=С* ; N7b< A  :Op= '#P KZ &)* PQ 4_5RQ-t[PPxL)KK ݥ 8 ; KΞ E#C }RPMHkrjwwwXeroxResearch RGBLinearG0i lP  ,iN(pIG/p ;6U! 4 ;h / O=~cVR0B  a X7 #/ WX?ը?Pg, A0h0+c 98^ xe]ykG0ikrjwwwXeroxResearch RGBLinearFeLNRH< Op@YM`)8?2$`27%5`6 "P;( X>{%4 J=x(ء~;"IsT#+;',[7m "Zp:!X@aRQ](D0+ j FekrjXeroxResearch RGBLinear0h}>o  :17"Xc y%$ x=|. ,23ء1=uY;3(XQ5E7 _%37 w/I83ءz) ) "X5衒"3W#L''xyx15'+I{p2kR1()(qxmw   pp //9\"8P;vZ0hkrjXeroxResearch RGBLinear&@ tQ &{a?R)$ Cֈ #W> $ ")[BPŰKD=_"Z BPo(8  u$AHt  i?AHС((Fp8P@L. ANTV= a," чT+'] T*Q5 *PV^];cFu !0 Wk'0g#I\, R .X/ /h@m 7 4qa'8b@  J8a8z"    &@krjwwwXeroxResearch RGBLinear2xu 2o8 ) h^k( aXPxW\ koWUBA| , Cvڧ1>@8hL3A'Y&/ 9' GN4 |xt3g;028 +n ,#X\y,f V%*P("[ >B @+KVP[:kO)Zpzq" YY1',mdWX uoK׻7`+@5  + !UHc^ E \Г IL+*, c  A 2xkrjXeroxResearch RGBLinearFp1( Mv a* RFg7 ,#XGI Im _o8@ pVk|(z'88%wȡ ȍ@WpTI4 c ;coN,`Сw)/x NR4H-{8Gx"P?4D`#u*)L[u8;%#6 +ܑVF X(_#5/C#JH`C.,%-P+\P &! Fp11krjXeroxResearch RGBLinear2Йl [2J.Xqr!:NR0D &V&ȡ< #&31^3Oh } 9  Nh Y5#!"J KP.JG"e(#X0($u %#i.F] H8>݊ \ W#;LGa)M@6kQ {*? `'x! U!h ߉>@03M;(%(j 0 4+X1>+Pk L%hR()03:G-G;QL 2СkrjwwwXeroxResearch RGBLinearLX`I2x#F>Fd_o\V1%Fj@0#"ʩr )Has 9VSSpQYh++1#A0LX`krjXeroxResearch RGBLinear?)  К m(@i ӣ - ߕWl83) 306 KK5 CHy uz`8E  oUrSB`hf hH 9$h !! h9j  " #>@<$ 9 g: hp"s 78 , -#an d/ X g+`  Y#]@.p h S W]  ӄc+f3!zI0eL?2# =(bD lo @h Lz @ A.  u! DcH O" #??krjXeroxResearch RGBLinear5n \{[ & h0p!LU%e02';(?) [vd7v Si#W2xbR07`FO# !PYKp #m%gGYh}cA/'([xW"Ae&1[x. 9b57AՊ 35nkrjXeroxResearch RGBLinearD? v  ء I880CQM&F>8y'%@m  ( I #;) U : 0B ( {A,08BHމ! e 7`Xb%4t:esxkrjXeroxResearch RGBLinear4Xw" mf]lXb=0iPf,T Ifh@IZ]0; bH@eZ+uh Y n \HIs@Y=x]LQ WlOX$+\;h":>  a@ Hi Qi !-eH h's2` a{mHpȵ!zp`@衒mZMN)5@aOO{,4Xkkgtop==Utop))KRuleRuletopIblockvv R  I reference6HY=UY>%wY[a Y%++{Yj2Y'&^Y 3Y]tY)VYQ` YN'YPDMsZG