FR2876818A1 - Vectorial format image file e.g. DXF format file, transcoding method for e.g. Wi-Fi wireless local area network, involves replacing image encoding instructions defining segments by single instruction defining polyline - Google Patents

Abstract

The method involves modifying image encoding instructions for associating an end of one segment of an image to another segment of the image, when distance between the end and the latter segment is less than a pre-established threshold to create a polyline. Number of instructions contained in a file is reduced by replacing the instructions defining the segments by a single instruction defining the created polyline. Independent claims are also included for the following: (A) a computer program comprising instructions for execution of a vectorial format image file transcoding method (B) an information storage medium comprising instructions for execution of a vectorial format image file transcoding method (C) an image file transcoder implementing a Vectorial format image file transcoding method.

Description

The present invention relates to a transcoding method and program

  an image file in vector format, a recording medium and a transcoder adapted for this method.

  A vector format is a format in which the image is encoded as a multitude of line segments, with each line segment corresponding to an elementary line of the image. These line segments are defined, for example, by the coordinates of their two ends, or by the coordinates of only one of these ends, a vector and a length. Known vector formats are, for example, DXF (AUTOCAD FILE) or SVG (SCALABLE VECTOR GRAPHIC).

  The vector format is particularly well suited to industrial design. Images in this format are therefore often generated from industrial design software such as AUTOCAD.

  However, industrial design software does not optimize the number of instructions needed to code an image in this vector format.

  The invention aims to overcome this drawback by proposing an image coding method for optimizing the number of instructions necessary to code this image in a vector format.

  The subject of the invention is therefore a method of encoding an image in a vector format, comprising: a step of modifying said instructions for connecting at least one end of a first segment of the image to a second segment of the image; same image when the distance between this end and the second segment is less than a first threshold pre-established so as to create a polyline, - a step of reducing the number of instructions contained in the file, by replacing the instructions defining the first and second segments by a single instruction defining the created polyline.

  The modification step makes it possible to connect a large number of segments that are not connected to one another so as to form polylines. In the field of industrial design software, the term polyline designates a geometric entity consisting of a succession of straight segments perfectly connected to each other by their ends. The coding of a polyline is carried out using a number of instructions strictly less than that necessary to code the same number of unconnected segments. The modified image is thus coded using a number of instructions less than that necessary to code the unmodified image. The amount of memory needed to record the coded picture is therefore reduced. In addition, the quality of the modified image is only slightly altered by the modification step since the segments that are connected to each other often correspond to segments that were not connected by mistake during the modification. realization of the image.

  Embodiments of this method may include one or more of the following features: - the modifying step consists of moving the end of at least one of the two segments so that this displaced end is merged with a point of intersection of the lines carrying the two segments; a step of merging segments consisting of replacing the instructions defining a third and fourth segment of the image by a single instruction defining a fifth segment of module and direction equivalent to the meeting of the third and fourth segments, when in addition: the angle between the lines carrying the third and fourth segments is less than a second predetermined threshold, the distance between one end of the third segment and its projection point on the line carrying the fourth segment is less than a third threshold pre-established, and, - the projection point belongs to the fourth segment or is located at a distance from one of the ends of the fourth segment lower than a fourth pre-established threshold; the reduction step comprises a search operation of the largest polyline before replacing the instructions defining the segments of this largest polyline by a single instruction defining this polyline, the magnitude of a polyline being measured by the number of segments that compose it; the reduction step comprises a construction operation of a tree structure in which each node corresponds to a segment and each link between two nodes indicates that the segments corresponding to these two nodes are connected to each other, and the operation of The search also involves finding the deepest path in this tree structure.

  The preceding features allow each in isolation to further limit the number of instructions necessary to encode an image in a vector format.

  The subject of the invention is also a computer program and an information recording medium comprising instructions for the execution of a coding method such as that described above when said instructions are executed by an electronic computer. .

  The invention also relates to a transcoder of an image file in a vector format adapted to implement a transcoding method as described above. The file has several instructions each defining a line segment, each segment having two ends. The transcoder comprises: a module for modifying said instructions for connecting at least one end of a first segment to a second segment of the image when the distance between this end and the second segment is less than a first predetermined threshold of to create a polyline, and - a module for reducing the number of instructions in the file, to replace the instructions defining the first and second segments by a single instruction defining the polyline created.

  The invention will be better understood on reading the description which follows, given solely by way of example and with reference to the drawings in which: FIG. 1 is a schematic illustration of the architecture of a transcoder of image files in a vector format; FIGS. 2A and 2B are an illustration of a flowchart of a method of transcoding an image into a vector format, and FIGS. 3, 4A, 4B, 5A, 5B, 6A and 6B are illustrations of segments of an image before and after execution of the transcoding process.

  FIG. 1 represents a transcoder 2 of image files in a vector format, connected to an information recording medium 4.

  The transcoder 2 comprises a module 6 for acquiring a file of an initial image in a vector format.

  This module 6 is, for example, capable of acquiring the file of an image generated by an industrial drawing software or of extracting such a file in vector format from a data source located on the Internet network. Here, the module 6 is able to acquire a file in SVG format.

  An SVG file has <line> instructions defining visible line segments in the image and optionally <polyline> instructions defining visible polylines in the image.

  An output of the acquisition module 6 is connected to an input of a file analyzer 12.

  The analyzer 12 is able to replace each <polyline> instruction of the acquired file with a succession of <line> instructions each defining a respective segment of the polyline. The sequence of instructions <line> is, for example, written in SVG format.

  An output of the analyzer 12 is connected to an input of a fusion module 14. This module 14 is able to replace two instructions <line> of the file analyzed by the analyzer 12, by a single instruction <line> defining a segment of module and direction equivalent to the meeting of the two segments replaced.

  An output of the module 14 is connected to an input of a segment connection module 16. The module 16 is able to modify the initial image by connecting the end of one segment to another segment of the image when the distance between this end and this other segment is less than a pre-established threshold.

  An output of the module 16 is connected to an input of a module 18 for reducing the number of instructions necessary to code the image in a vector format. For this purpose, the module 18 comprises: - a submodule 20 for the construction of a tree structure representative of the connections between segments, - a sub-module 21 for searching the largest polyline, and - an encoder 22 specific to to code with the help of an instruction of the SVG language the polyline found by the sub-module 21.

  The modified file generated by the reduction module 18 is in SVG format.

  An output of the module 18 is connected to a conventional file compression module 30, either directly or via a format conversion module. Here, by way of illustration, the device 2 comprises a single format conversion module 32 capable of converting an SVG file into a DXF format file before transmitting it to the compression module 30.

  The module 30 is, for example, here a module capable of generating a compressed file in ZIP (PKzip file name extension) format.

  The transducer 2 is here made from a conventional programmable electronic computer capable of executing instructions recorded in the information recording medium 4. For this purpose, the information recording medium 4 comprises instructions for the execution of the method of FIG. 2 when said instructions are executed by the electronic computer.

  The operation of the transducer 2 will now be described using the flowchart of FIGS. 2A and 2B and with reference to FIGS. 3, 4A, 4B, 5A, 5B, 6A and 6B.

  In a step 50, the module 6 acquires a file in a vector format of an image. Here, the file is in SVG format. This file is then transmitted to the analyzer 12 which replaces, in a step 52, each <polyline> instruction defining a polyline by several <line> instructions each defining a single segment of the polyline. At the end of this step 52, the image encoded by the analyzed file is identical to that acquired during step 50.

  The analyzed file is transmitted to the module 14 which then proceeds to a step 54 of merging segments of the image.

  The objective of step 54 is to replace by a single <line> instruction two almost redundant <line> instructions. More precisely, during step 54, two <line> instructions defining segments S1 and S2 are replaced by a single <line> instruction defining a segment S3 of module and direction equivalent to the union of two segments S1 and S2 if the angle between the straight lines D1 and D2 carrying the two segments S1 and S2 is less than a pre-established threshold dc, and the distance between one end of the segment S1 or S2 is its projection point P'i on the right bearing the other segment is less than a pre-established threshold dd, and - the projection point P'i belongs to the other segment or is at a distance from the ends of the other segment less than a preset threshold dt .

  The operations of step 54 will be described in the particular case of the two segments S1 and S2 of FIG.

  The segment S1 has two ends P1 and P2 while the segment S2 has two ends P3 and P4. The segment S1 is carried by the line D1 and the segment S2 is carried by the line D2.

  Step 54 begins with an operation 56 for calculating a director coefficient Ci for each of the segments defined in the analyzed file. A steering coefficient is representative of the inclination of the segment relative to a reference direction. For example, the director coefficient Ci of a segment Si defined by two end points P1 and P2 is calculated using the following relation.

  YP2 - YP1 Ci = xP2 - xP1 where: - Ci is the directional coefficient of the segment Si, - xp1 and ypl are respectively the abscissa and the ordinate of the end P1 in an orthonormal frame, and - xP2 and yP2 are respectively the abscissa and the ordinate of the end P2 in the same orthonormal coordinate system.

  Then, the directional coefficients of the segments are compared two by two during an operation 58 to select only pairs of segments having directions close to each other. The directions of two segments are close if the difference between the coefficients of the two segments is less than the preset threshold dc. For example, in the particular case of FIG. 3, the directing coefficient C1 of the segment S1 is subtracted from the directing coefficient C2 of the segment S2. If the difference in absolute value between C1 and C2 is lower than that of then the module 16 carries out an operation 60.

  During the operation 60, the module 14 calculates the projection of the ends of a first segment of the pair on the line carrying the second segment. The projection is made in a direction orthogonal to the line carrying the first segment.

  In the case of FIG. 3, the module 14 calculates the projection of the ends P1 and P2, denoted respectively P'1 and P'2, on the line D2.

  Then, during a step 62, the distance separating the end of the segment from its calculated projection point is calculated. This distance is noted dl in Figure 3 for the end P1 and d2 for the end P2.

  Then, during an operation 64, the distances calculated during the operation 62 are compared with the predetermined threshold dd so as to select only segments that are close to one another. In the case of Figure 3, it is assumed that the distance d2 is below the threshold dd.

  If so, that is to say if one of the distances calculated during the operation 62 is below the threshold dd, then the module 14 checks, during an operation 66, if at least one projection points of the first segment belong to the second segment. For example, in the case of FIG. 3, the module 14 checks whether at least one of the points P'1 and P'2 belongs to the segment S2. In FIG. 3, the point P'2 belongs to the segment S2, which is not the case for the point P'1.

  If so, the module 16 replaces, in an operation 68, the <line> instructions defining the two segments S1 and S2 by a single <line> instruction defining the segment S3.

  In the opposite case, if during operation 66, none of the projection points belong to the second segment, then the module 14 determines, during an operation 70, whether the distance between one of the protection points and one of the ends of the second segment is below the pre-established threshold dt. This pre-established threshold is here chosen equal to dd. If so, the module 16 proceeds to the replacement operation 68.

  During operation 68, the module 14 selects, during an operation 74, the two ends farthest from each other in the set {P1; P2; P3; P4} ends of the two segments S. and S2. Then, during a sub-operation 76, it replaces the <line> instructions defining the two segments S1 and S2 by a single <line> instruction defining the segment S3 having for its ends those selected during the sub-operation 74. new segment created previously is reintroduced in the list of segments to be processed.

  Upon completion of operation 68, the process returns to operation 58 and reiterates operations 58 to 68 for a new pair of segments. These operations are repeated as long as there are untreated pairs of segments in the analyzed file.

  If during operation 58, the compared key coefficients are greater than the threshold d, or if during the operation 64 the calculated distances are greater than the threshold dd or if during the operation 70 none of the ends of the second segment n is close to one of the projection points, so the process also returns to operation 58.

  Thus, step 54 reduces the number of instructions necessary to code the image while limiting the changes made to the initial image by the thresholds of, dd and dt.

  Then, the file processed by the module 14 is transmitted to the module 16. The module 16 modifies, during a step 80, the coded image by connecting the ends of the secant segments, that is to say the segments whose difference of the guiding coefficients and greater than the threshold dc.

  This step 80 aims to modify the <line> instructions of the merged file to connect the segments to each other and thus create polylines instead of a succession of unconnected segments.

  Three different cases of secant segments are shown in Figures 4A, 5A and 6A. Figures 4B, 5B and 6B respectively illustrate the segments of Figures 4A, 5A and 6A after the modifying step has been performed. The notations used for FIGS. 4A, 4B, 5A, 5B, 6A and 6B are identical to those defined with reference to FIG. 3. It will be noted here simply that the point of intersection between the lines each bearing one of the segments secants is noted P '.

  Initially, during an operation 82, the module 165 selects two <line> instructions defining the segments S1 and S2 in the merged file.

  Then, during an operation 84, it calculates the coordinates of the point of intersection P 'between the straight lines D1 and D2 bearing the segments S1 and S2.

  Afterwards, it calculates, during an operation 86, the distances L1, L2, L3 and L4 separating the point of intersection P 'respectively of the ends P1, P2, P3 and P4.

  Then, during an operation 88, the module 16 determines whether the intersection point P 'belongs to the segment S1 and / or the segment S2 (case of FIGS. 4A and 5A). If so, the module 16 compares, during an operation 90, the distances L1, L2, L3 and L4 to a preset threshold Di.

  If one of these distances is less than the threshold Ai, then the module 16 carries out an operation 92 of moving the end or ends closest to P 'to the point P'. More specifically, during the operation 92, at least one of the <line> instructions defining one of the segments is modified to define a new segment S'i whose end has been moved to connect this segment S 'i to the other segment.

  For example, in the case of FIG. 4A, the distance L2 separating P2 from the point P 'is assumed to be smaller than Di and the end P2 is displaced to the point P' so that the modified segment S'1 is connected to S2 segments as shown in Figure 4B.

  In the case of Fig. 5A, both the distances d2 and d4 are below the threshold A1 so that the ends P2 and P4 are both displaced to the point P '. The result of this change is shown in Figure 5B.

  If during the operation 88 it is determined that the point P 'does not belong to the segments S1 and S2, (case of FIG. 6A), then the module 16 carries out an operation 94 for testing the proximity of the ends of the segments.

  During the operation 94, the module 16 determines if one of the two distances L1 and L2 is less than a pre-established threshold Ae and if one of the two distances L3 and L4 is less than this same pre-established threshold If so, the nearest end of the segment S1 and the end closest to the segment S2 are moved to the point P 'during an operation 96. This operation 96 is performed, as in the case of operation 92, modifying the <line> instructions defining these segments S1 and S2. This is shown in Figure 6B.

  After the operation 92 or 96 or if during the operation 90 it is determined that none of the distance Li is less than the threshold Di, or if during the operation 94 it is determined that the two distances L1 and L2 or L3 and L4 are greater than threshold Ae, then module 16 returns to step 88 and executes operations 82 through 96 again for a new pair of segments.

  Operations 88 to 96 are repeated until all the pairs of segments have been processed.

  At this point, the file modified by step 80 contains the same number of <line> instructions as the file merged by the module 14.

  The modified file is then transmitted to the module 18 to reduce, during a step 100, the number of <line> instructions contained in the modified file.

  During step 100, the sub-module 20 constructs, during an operation 104, a tree structure representing the connections of the segments between them. In this tree structure, each segment is represented by a node and the two nodes are interconnected if the corresponding segments are connected to each other.

  Then, during an operation 106, the submodule 21 searches for the deepest path in the tree structure constructed during the operation 104. The deepest path is the one with the largest number of nodes and corresponds to the largest polyline, that is, the polyline with the largest number of segments. For this purpose, search algorithms for the deepest conventional path are implemented, such as the Dijkstra algorithm.

  Once the largest polyline has been found the coder 22 encodes, in an operation 108, the polyline found using a single coding instruction. More precisely, during this operation 108, the <line> instructions defining in the modified file the segments belonging to the found polyline are replaced by a single instruction defining the polyline found.

  For example, in the case of the SVG format, the instruction used will be a <polyline> instruction.

  Once the operation 108 is complete, the process returns to the operation 104. The operations 104 and 108 are repeated until there is no more polyline to code in the modified file.

  This way of proceeding to reduce the number of <line> instructions of the modified file by first coding the largest polylines makes it possible to optimize the number of instructions necessary to code the image.

  At the end of step 100, during a step 120, if the output format of the reduced file is not the SVG format, then the file reduced by the module 18 is transmitted to the module 32 of conversion of format. The module 32 converts, during a step 122, the reduced file to the SVG format in another format such as the DXF format.

  At the end of step 122 or in the case where the output format is the SVG format, the reduced file is transmitted to the compression module which compresses, in a step 124, this file using conventional algorithms.

  The different pre-established thresholds used in the device 2 limit the importance of the modifications made to the coded image. They are chosen to constitute a good compromise between the importance of the modifications made to the coded image and the number of instructions necessary to code it. Indeed, the higher these pre-established thresholds, the less the number of instructions needed to code the image is important and the visual appearance of the image may be changed.

  The device and method described herein are particularly effective for encoding images generated by industrial design software. In particular, the module 14 reduces the number of instructions while the module 16 modifies the image to create polylines instead of unconnected segments. Each polyline is then encoded using a single instruction which ultimately reduces the number of instructions needed to encode the initial image.

  Such a coding method is of particular interest for the transmission over a network of images in vector format. This network is a WiFi network or a wireless communication network type UMTS (Universal Mobile Telecommunication System).

  The device 2 and the method of FIG. 2 also make it possible to reduce the memory footprint of the coded image, which makes it possible to display such images on terminals equipped with a restricted memory such as, for example, a mobile telephone.

  Here, the image is modified during step 80 by moving the end of at least one segment so that it is coincident with the point of intersection of the lines carrying the two segments. This makes it possible to limit the number of points of the polyline created and thus the size of the file containing the instructions coding the modified image. However, alternatively, the connection of two segments whose ends are close and realized by adding a <line> instruction defining a new segment connecting the two ends close to each other. Thus, in this variant, the ends of the segments are not displaced but connected to each other via a new segment. This also makes it possible to create polylines and thus to reduce the number of instructions necessary to code the image.

Claims (1)

16 CLAIMS   A method of transcoding an image file into a vector format, the file comprising several instructions each defining a line segment in the image, each segment having two ends, characterized in that the method comprises: a step (80) modifying said instructions for connecting at least one end of a first segment of the image to a second segment of the same image when the distance between that end and the second segment is less than a first predetermined threshold of to create a polyline, - a step (100) of reducing the number of instructions contained in the file, replacing the instructions defining the first and second segments by a single instruction defining the polyline created.   2. Method according to claim 1, characterized in that the modifying step (80) consists in moving the end of at least one of the two segments so that this displaced end is coincident with a point of intersection of straight bearing both segments.   3. Method according to claim 1 or 2, characterized in that it comprises a step (54) of merging segments consisting of replacing the instructions defining a third and fourth segments of the image by a single instruction defining a fifth segment. of modulus and direction equivalent to the meeting of the third and fourth segments, when in addition: - the angle between the straight lines carrying the third and fourth segments is less than a second predetermined threshold, 2876818 17 - the distance between one end the third segment and its point of projection on the straight line bearing the fourth segment is less than a third pre-established threshold, and, - the projection point belongs to the fourth segment or is at a distance from one of the ends of the fourth segment less than a fourth pre-established threshold.   4. Method according to claim 3, characterized in that the reduction step (100) comprises an operation (106) for finding the largest polyline before replacing the instructions defining the segments of this largest polyline with a only instruction defining this polyline, the magnitude of a polyline being measured by the number of segments that compose it.   5. Method according to claim 4, characterized in that the reduction step comprises an operation (104) of construction of a tree structure in which each node corresponds to a segment and each link between two nodes indicates that the segments corresponding to these two nodes are connected to each other, and in that the search operation further consists of searching for the deepest path in this tree structure.   6. Computer program, characterized in that it comprises instructions for executing a transcoding method according to any one of the preceding claims, when said instructions are executed by an electronic computer.   7. An information recording medium, characterized in that it comprises instructions for the execution of a transcoding method according to any one of claims 1 to 5, when said instructions are executed by an electronic calculator .   8. Transcoder of an image file in a vector format adapted to implement a transcoding method according to any one of claims 1 to 5, the file comprising several instructions each defining a line segment, each segment having two ends, characterized in that the transcoder comprises.   a module (16) for modifying said instructions for connecting at least one end of a first segment to a second segment of the image when the distance between this end and the second segment is less than a first predetermined threshold; creating a polyline, and a module (18) for reducing the number of instructions in the file, capable of replacing the instructions defining the first and second segments with a single instruction defining the polyline created.