FR2700036A1 - Two=dimensional image storage and visualisation - Google Patents

Abstract

The process involves storing digital units of information in memory for the purpose of visualisation. Two dimensional images are made up of a number of pixels for high line definition to produce full screen definition. The definition is more than necessary for the full screen definition. The screen is a file of pixels which can represent "n" files of pixel/lines where each file is the base of "n" images and can be stored as compressed files. The size of the files varies from the principal file corresponding to the full screen definition required. After partition one of the "n" base files is integrated in one or several files corresponding to a lower definition from the main full file. The sub-files are produced by reduction steps calculated through a filtering method. Intermediate steps result in the production of a partitioned file to produce an image with "p" pixels less than "n".

Description

METHOD AND DEVICE FOR STORAGE AND VISUALIZATION
LARGE IMAGE IMAGE
The present invention relates to a method and a device for its implementation, intended to allow storage in memory on the one hand, and consultation by display on the screen, on the other hand, of two-dimensional graphic representations, such as large-sized images, said image being viewed in digital form with a definition of N pixels (points) x M lines, said definition being greater than the definition of viewing means, such as usual computer viewing screens which have a definition hereinafter called basic definition.

 The total amount of digital information corresponding to the whole of a large image is relatively high.

This number depends mainly on the definition (pixels x lines) desired by the user. The higher the definition, the greater the number of pixels multiplied by the number of lines. This therefore results in a very high number of information which it is necessary to store in large capacity memories, which entail relatively high investment and operating costs.

 Furthermore, the consultation of large images, and in particular the viewing of these on a computer-type screen is subject to specific constraints, such as for example the speed of access to information and the quality of the display, and also the number of options offered to the user, for example the number of details that can be viewed and the magnification ratio thereof.

 According to known techniques, the consultation of large documents such as images or graphic representations is carried out by reading the file containing the digital information relating to this image, according to a relatively high resolution, then an image is reconstituted by subsampling. of a size corresponding to the display screen, that is to say an image of the size of the screen. This subsampling is carried out by sampling a point on n (n> 2) as a function of the magnification ratio, similarly for the lines of which only one line on n is taken into account to constitute the image which will be displayed at the screen.

This process is very time-consuming, and it is not uncommon to wait for one or three minutes for information on the screen after the user's request. This duration is considered, in the field of data processing, as unacceptable and is in any event incompatible with economic imperatives.

 We have tried to remedy these drawbacks by producing in advance, subsets # of the image according to determined magnification ratios. However, on the one hand this requires large memory capacities, and on the other hand, this is not completely satisfactory for the user who is given magnification ratios which do not necessarily correspond to his wish, without counting that the access time to the images remains relatively long.

 Thus, there is a hindrance to consultation in the form of viewing on screen of large images, especially given the extremely long access time to a detail of the latter or to the entire image.

The present invention overcomes these drawbacks and proposes a method and a device for its implementation making it possible to store in memory, with relatively modest means, a large quantity of information, that is to say numerous images and big size. The method and the device of the invention also and above all allow the consultation by visualization of the images, of the details chosen of the latter, and / or any combinations of images between them, and this, with reduced access times compared in the prior art, up to a factor of 10 or more.
To this end, according to the invention, the method for storing in memory and consulting by display means, a digital information unit corresponding to a two-dimensional graphic representation (image), said unit containing a number of pixels and of line greater than the definition (N pixels x M lines) of said display means (called full screen definition), characterized in that:
a) the said image is digitized according to the maximum definition desired (greater than the full screen definition3 then a raw file is obtained called the starting file;
b) partitioning said initial file into “n” base files each comprising a number of pixels / lines equal to the full screen definition, each base file containing a detail of the image;
c) each of the "n" basic files is subjected to compression then said files are stored;
d) the size of the starting file is reduced so as to obtain a main file corresponding to said image and to the full screen definition. The possible remainder of the starting file, after partitioning, and which is not part of one of the " n "basic files, is integrated into one or more annex files corresponding to a definition lower than the maximum definition.

 The reduction step is preferably carried out by calculation, for example, by filtering by means.

 Advantageously, the reduction includes at least one intermediate step resulting in the creation of "p" intermediate files, with "p" less than "n", all of said "p" files corresponding to said image with an intermediate definition between the definition maximum and full screen definition, said "p" intermediate files being preferably subjected to compression.

 According to an advantageous embodiment, the resolution in N pixels x M lines of basic files corresponds to a digital television standard or standard (720 pixels x 576 lines), and the maximum definition is 2n N pixels x 2n M lines, M being an integer greater than or equal to one and preferably equal to 2 (the maximum definition then being 2880 pixels x 2304 lines, screen of the 4/3 type).

 Preferably, the method further comprises a step of reducing the size of the main file so as to obtain at least one reduced file, corresponding to said image of reduced size, and to a definition lower than the basic definition (N x M ), said reduced definition being equal to N / mx M / m. Preferably, m = n.

 The invention is likely to find many applications, and in particular to name only some of them, in museums, any organization having numerous archives (photo libraries, etc.), universities, insurance companies, libraries, imagery documents. medical, research and heritage management centers.

 The method according to the invention can be implemented using a memory unit and a consultation station, or a central memory unit associated with different consultation stations arranged in a network, or alternatively '' a central archiving unit capable of being interrogated remotely by one or more consultation / viewing posts.

The invention will be clearly understood in the light of the example of the implementation of the invention described below with reference to the accompanying drawings in which:
- Figure 1 shows five graphic representations, the size of a display screen, and each showing an image at different scales and resolutions;
- Figure 2A shows the different levels of detail and enlargement, and resolution arranged in a pyramid, on a scale to facilitate understanding;
- Figure 2B shows the pyramid of Figure 2A on a real scale;
- Figure 3 shows five screens representing the same image and showing the scale relationship from one screen to another;
- Figure 2B shows the pyramid of Figure 2A on a real scale;
- Figure 4 shows a schematic block diagram of the method according to the invention and the means for its implementation; and
- Figure 5 shows schematically two sets of high resolution files in partial recovery.

 FIG. 1 shows a series of rectangles symbolizing a display screen. The five rectangles are of equal size and correspond for example to a 4/3 type format. Five successive rectangles 1A, 1B, 1C, 1D and 1E are represented. Each rectangle, corresponding to a screen, comprises an image or series of images or details of this same image. Thus, the first rectangle tA comprises sixteen images, in this case in the example represented sixteen times the same image, namely a character on a balance bike. The second rectangle 1B comprises four times the same image, on a larger scale, while the screen 1 C comprises the entire image of the character and the balance bike. The screen 1C is called the main screen. The screen 1D represents a part or detail of the image represented on the main screen 1C, namely the bust and the head of the character. Finally, the last screen 1E represents a very enlarged detail of the balance bike.

The main screen 1C shows the entire image for a given definition of N pixels x M lines. For example, the definition could advantageously be 720 pixels or points x 576 lines. This definition corresponds to the definition of digital television defined by the CCIR (International Consultative Committee of
Broadcasting). This standard was chosen as a reference in the method according to the invention, but it is understood that any other standard could be suitable. The screens tA and 1B show several images, on a reduced scale, and therefore at a lower definition.

Conversely, screens 1D and 1E show details of the image, on an enlarged scale, but at a higher definition than that of the main screen 1C. For example, the screen 1A shows sixteen images or thumbnails whose definition of each is 144 x 180, while the four thumbnails shown on the screen 1B correspond to a definition of 288 x 360, this without a main screen 1C showing the image in full screen at a resolution of 720 x 576.

The 1D screen shows a detail of the image at a resolution of 1440 x
1152, while the enlarged detail shown on the screen 1E corresponds to a definition of 2.880 x 2.304.

 With reference to FIG. 2A, the method of the invention, from a large image, that is to say one whose definition in number of pixels and of lines is greater or much greater than the definition of usual visualization means such as a computer screen, an information unit is produced for this image, by digitizing the latter according to a maximum definition desired for the user. In the example shown, this maximum definition is 2,880 pixels x 2,304 lines.

 In a second step, we divide or partition this unit of information corresponding to a starting file containing the pixels and the lines, into a number "n" of basic files each comprising a number of pixels / lines substantially equal to the screen definition. Each basic file therefore corresponds to a detail of the complete image. On the pyramid, the meaning of which will be explained later, the base of the latter constitutes the complete image, for the maximum definition of 2.880 x 2.304. Each subunit is referenced by a capital letter, from H to W. In the example, the number "n" of subunits each corresponding to a file from H to W is 16. The base of the pyramid therefore has 4 x 4 files. Note that each basic file H to W presents a quantity of information for a definition corresponding to that of the screen, that is to say 720 x 576. Indeed, the definition of 2.880 x 2.304 corresponds to four times the definition of 720 x 576. Thus, in the example shown in FIGS. 1 and 2, the display screen 1E shows a detail of the image corresponding to one of the basic files H to W.

 Note that the successive levels of the pyramid in Figure 2A have been shown, for better understanding, equidistant from bottom to top. In reality, the levels are arranged as shown in FIG. 2B in order to respect the enlargement / reduction ratio of the example; let "l" be the length separating the levels H-W and D-G on the edge of the pyramid. The length between level D-G and level C will be 1/2 and so on.

Once the basic files H to W have been created, these files are subjected to a compression step, known in itself. Then the starting file is reduced (by half in the example shown) by
so as to create an intermediate file, the latter then being partitioned into four files identified by the letters D, E, F and G and constituting together an intermediate level of the pyramid, arranged at a level higher than the basic level constituted of the sixteen basic files H to W. Each intermediate file DG corresponds to the definition of the screen. The total image changes from the definition of 2.880 x 2.304 and is thus reduced to a definition of 1.440 x 1.152. With reference to FIG. 1, the files D to
G contain information in the form of a detail of the image, as shown for example on the screen 1D in FIG. 1. The files D to
G are then themselves subjected to compression.

 The reduction is continued, from the intermediate file that serves to create the files From To, until a main file bearing the reference C and containing an amount of information corresponding to the representation of the full image on the screen, as represented on the image is obtained. main screen 1C of figure 1.

The definition corresponding to the main file C is 720 x 576.

 Advantageously, the reduction is continued until a file B is obtained, the amount of information of which corresponds to the representation of the complete image, on a reduced scale, and for a definition lower than that of the main file C, namely 360 x 288. Thus, the file B comprises the image on a scale as represented on the screen 1B of FIG. 1. In the example given, the screen 1B shows four images on a reduced scale or thumbnails.

 We can also continue the reduction of file B until a file A of lower definition is obtained, for example 180 x 144 and corresponding to the representation of the image in the form of a thumbnail on a smaller scale than that shown on screen 1B. The thumbnails corresponding to the files A are of size as shown on the screen 1A. The reduced scale allows viewing of sixteen thumbnails. Screens 1A and 1B, although showing the same image several times, can represent images of the same scale, on the same screen, but different from one to another, which allows the user to choose the image that interests him among many.

 It is understood that the exemplary embodiment of the pyramid shown in FIGS. 2A and 2B is only a preferred embodiment and that the number of stages of the pyramid varies in particular as a function of the maximum definition desired by l user (files H to W) and according to the dimensions of the image to be viewed.

Thus, for relatively small or medium-sized images, it is entirely possible to digitize the total image according to a definition less than 2,880 x 2,304, and for example corresponding to 1,440 x 1,152. Thus, the reduction of the file containing the files D to G thus obtained by the partition of the starting file makes it possible to directly lead to the creation of the main file C corresponding to the full screen image (screen 1C of FIG. 1). Likewise, the reduction ratio from one level to the next, equal to 4 (on the surface) in the example of FIGS. 2A and 2B can be modified as required.

Also, the pyramid possibly contains levels ineluting additional files resulting from the rest of the partition into "n" basic files of the initial file.

 In the example shown, the process results in the creation of twenty-three files, sixteen basic files H to W, four intermediate files D to G, a main file C, and two files A and B, corresponding to lower resolutions .

 It should be noted that the reduction steps mentioned above are preferably carried out with algorithms making it possible to avoid the sub-sampling artifacts. Also, compression will preferably be carried out via the JPEG standard.

 FIG. 3 shows the possibilities offered to the user who can thus choose from the screen 3A, representing sixteen thumbnails, for example of different images, the image which interests him.

Also, he can choose a set of four different images on an enlarged scale, as shown on the screen 3B. Once his choice has been made on an image, he can view it in full screen, as shown on the 3C screen. The passage from one visualization to another, shown on the successive screens 3A, 3B and 3C is done by calling the corresponding files in the pyramid shown in FIG. 2.

Thus, screen 3A shows a set of files of type A, while screen 3B shows four files of type B, and screen 3C shows a main full image file referenced C. On screen 3C has been represented a rectangle 3F corresponding to a detail likely to interest the user who, to view this detail, calls the corresponding file (one of the files D to G in FIG. 2). It is possible to further increase the magnification and benefit from a detail view on a larger scale, with a higher resolution such as that surrounded by the 3G rectangle on the 3C screen. This detail is shown to a larger scale and definition on the 3E screen.

The visualization of this detail on a scale corresponding to a very high definition (for example of 2.880 x 2.304) is done by calling from one to four basic files H to W of figure 2. In the example represented, the size of the screen is such that any window (screen size) intercepts four basic files (H to W).

 It is understood that the method and the device of the invention allow rapid access to any detail view likely to interest the user and this, to an appropriate definition according to the needs of the user. Also, the user can choose from a very large number of images represented in the form of thumbnails on a reduced scale (screen 1A or 3A).

In Figure 4, there is shown schematically a block diagram of the method of the invention and the device for its implementation. Starting from the image referenced 10, of large dimension, the latter is digitized using known means such as a digitizing camera bearing the reference 11 capable of scanning the image according to columns 12 and rows 13. The lines 13 carry a certain number of pixels 14. The data from the camera 11 are stored in the memory 15 in the form of a starting file. Means 16 known in themselves make it possible to divide or partition all of the information made up of lines and pixels corresponding to image 10, according to "n" files, "n" being for example equal to 16. by camera
It has been produced according to a definition corresponding to the maximum definition desired by the user, for example 2.880 x 2.304.

 The partition carried out by the means 16 leads to the creation of 16 sub-units or basic files (corresponding to the files H to W of FIG. 2). Screens 1E and 3E show an example of such a file H to W. These files are subjected to compression, according to the JPEG standard, for example, using compression means 17. These basic compressed files H to W are sent to a main memory 18. The initial file (in memory 15) is reduced 1/4 (block 19) to give an intermediate file, then subjected to a partition (block 20) in basic DG files themselves tablets (block 21). The D-G files are then stored (18). The intermediate file resulting from the reduction 1/4 of block 19, is subjected to a new reduction (block 22) of 1/4 in order to obtain a main file C. The main file C is itself compressed (23) then stored in the main memory 18. The main file resulting from the reduction 22 is routed to reduction means 24 allowing the creation of a file B itself on the one hand subjected to compression 25, and stored in the main memory 18, and on the other hand, routed to reduction means 26, themselves resulting in the creation of a file A which is stored after compression 27 in the main memory 18. The small files A and B occupy less than the screen, they are no longer partitioned because they are contained in the full screen definition. The main memory 18 thus contains 23 files, namely 16 basic files H to W, 4 intermediate files D to G, a main file C and two files A and B. All of these files are accessible for viewing by the intermediate of means 23, with a view to viewing on a screen 28 using a control station 29 (keyboard or mouse) of an image or of certain parts of the image as mentioned above, with reference to Figure 3.

 It is possible to process images of greater or lesser size, and the processing of very large images can be carried out by scanning the latter according to a very high or ultra-high definition (containing a very large number of pixels and a very large number of lines). In summary, the size of the initial file is reduced, for example to half of pixels and half of lines, in successive reduction steps until a main file C (full screen) is obtained. In the example given, the passage from the initial file (maximum definition) to the main file (full screen definition) is carried out in a single intermediate operation.

 It is described below, a variant allowing the processing of very large images by using a set of relatively small basic files, for example 16 in the example described and mentioned above. FIG. 5 shows a first group 30 of 16 basic files, corresponding to the base of a first pyramid, while a second group 31 of 16 other basic files corresponds to another pyramid. The 16 files in group 30 and the 16 files in group 31 are partially overlapped by three files from each of the two pyramids. The hatched area represents the image to be displayed or an area of particular interest to the user which itself is large and which nevertheless must be viewed at a very high or ultra-high definition. shown schematically in Figure 5 shows the hatched area 32 which is likely to be thus viewed at a very high definition, although being very large and in any case, greater than the base of the pyramid formed by the 16 files basic. The user who has called one of the basic files containing part of the hatched area can thus first call the basic files of the second pyramid 31 and then by calling neighboring files, end up with one of the common files between pyramids 30 and 31 and thus "pass" from one pyramid to another, that is to say from a first group of basic files to the second group of basic files and thus continue the exploration and analysis of the hatched area to be stigmatized.

 The method according to the invention allows the constitution of a database from images, and the consultation of these large images, in a simple and above all rapid manner. Note also that the use of standards allows the evolution and adaptation of the method of the device according to the invention to all existing systems.

 Also, the compression of the files makes it possible to substantially reduce the memory space intended for the storage of all the files (for example of the 23 files of the pyramid of FIG. 2). The memory space saving ratio can be of the order of 40. In fact, compression makes it possible to achieve gains in a ratio of the order of 1 to 10, a screen makes it possible to represent a quantity of information relating to four basic files H to W.

 Also, it is possible to implement the method according to the invention remotely, that is to say that the main memory 18 containing the digitized and compressed files can be placed in a location remote from the consultation and viewing center. .

Claims (12)

 1 - Method for storing in memory and consulting by display means, a digital information unit corresponding to a two-dimensional graphic representation (image), said unit comprising a number of pixels and lines greater than the definition (Npixels x M lines) of said display means (called full screen definition),  characterized in that:  a) said image is digitized according to the maximum definition desired (greater than the full screen definition); we then obtain a file called the start file;  b) partitioning said initial file into “n” base files each comprising a number of pixels / lines equal to the full screen definition, each base file corresponding to a detail of said image;  c) each of the "n" basic files is subjected to compression then said files are stored; and  d) the size of the initial file is reduced so as to obtain a main file corresponding to said image and to the full screen definition.  2 - Method according to claim 1, characterized in that the rest of the starting file, after partition, and which is not part of one of the "n" basic files, is integrated into one or more annex files corresponding to a definition lower than the maximum definition.  3 - Method according to one of the preceding claims, characterized in that said reduction step is carried out by calculation, by filtering by average.  4 - Method according to one of the preceding claims, characterized in that said reduction step comprises at least one intermediate step, resulting in the creation of an intermediate file, the latter then being subjected to a partition in "p" files, with "p" less than "n", all of said "p" files corresponding to said image with an intermediate definition between said maximum definition and said full screen definition.  5 - Method according to claim 4, characterized in that said "p" intermediate files are subjected to compression and then stored.  6 - Method according to one of the preceding claims, characterized in that the resolution in N pixels x M lines of each basic file corresponds to a digital television standard or standard (720 pixels x 576 lines).  7 - Method according to one of the preceding claims, characterized in that said maximum definition is 2n N pixels x 2n M lines, M being an integer greater than or equal to 1.  8 - Process according to claim 7, characterized in that M is equal to 2 and the maximum definition is 2.880 pixels x 2.304 lines, for display means allowing a representation in 4/3 proportions.  9 - Method according to one of the preceding claims, characterized in that it further comprises a step of reducing the size of the main file so as to obtain at least one reduced file, corresponding to said image of reduced dimension, and to a definition lower than the full screen definition (N x M), said reduced definition being equal to N / mx M / m.  10 - Process according to claim 9, characterized in that m = n.  11 - Device for storage in memory and consultation by display means, of a digital information unit corresponding to a two-dimensional graphic representation (image), said unit containing a number of pixels and lines greater than the definition (N pixels x M lines) of said display means (called full screen definition), characterized in that it comprises:  - Means for digitizing said image according to the maximum definition desired (greater than the full screen definition) and means for creating from this digitization a file called starting file;  - Means for partitioning said initial file into "n" basic files each comprising a number of pixels / lines equal to the full screen definition, each basic file corresponding to a detail of said image;  - means for compressing said "n" base files, and means for storing these; and  means for reducing the size of said initial file, so as to obtain a main file corresponding to said image and to the full screen definition.  12 - Device according to claim 11, characterized in that it further comprises means making it possible to reduce the size of the main file so as to obtain at least one reduced file, corresponding to said image of reduced dimension, and to a definition lower than the full screen definition (N x M), said reduced definition being equal to N / mx M / m.