FR2866997A1 - Digital image encoding method, involves choosing same values for quantization steps, and comparing steps with predetermined threshold, for set of frequencies lower than predetermined frequencies - Google Patents

Abstract

The method involves choosing same values for quantization steps. The steps are compared with a predetermined threshold, for a set of frequencies of current block, where the frequencies are lower than predetermined frequencies. The steps are replaced by the threshold to obtain a permanent quantization step, if the steps exceed the threshold. The permanent quantization step is stored in a quantization table. Independent claims are also included for the following: (A) a communication apparatus including an image encoding apparatus (B) an information storage unit readable by a computer or microprocessor, for executing an image coding method (C) a computer program product including instruction sequences for executing an image coding method.

Description

METHOD AND DEVICE FOR CODING AN IMAGE

  The present invention relates to a method and an encoder of an image.

  The invention belongs to the field of compression of digital images and, more generally, to the field of coding and compression of digital signals, whether video, audio, data, etc. There are various ways to compress lossy digital images. For example, US-A-5,883,979 discloses an image compression technique that uses three predetermined quantization values in three different areas of a JPEG quantization table. The ratio between these values is set by the user.

  However, this technique has the disadvantage of introducing into the images block effects when the compression ratio is high. These 15 block effects are very troublesome visually.

  The present invention aims to overcome this disadvantage.

  For this purpose, the present invention provides a method of coding an image divided into a plurality of data blocks, each block undergoing a frequency transformation, this method being remarkable in that it comprises the steps of: - choose a plurality of quantization step values; then, for a plurality of frequencies of the current block, comparing the current quantization step with a predetermined threshold; and if the current quantization step exceeds the threshold, replace the step with the threshold so as to obtain a definitive quantization step.

  Thus, the invention makes it possible to code a digital image with a high compression ratio without introducing a block effect, while remaining as close as possible to the optimal compression / distortion compromise.

  In addition, this technique is compatible with the JPEG standard and its execution is very simple and very fast.

  In a particular embodiment, the above-mentioned plurality of frequencies of the current block consists of a set of frequencies below a predetermined frequency.

  This makes it possible to affect a threshold only at the frequencies responsible for the block effects and, consequently, to overcome these block effects.

  In a particular embodiment, the method further comprises a step of storing the value of the definitive quantization step in a quantization table.

  In a particular embodiment, the blocks are 8x8 pixels in size.

  The two preceding particular embodiments make it possible to facilitate the compatibility of the method according to the present invention with the JPEG standard.

  In a particular embodiment, the step of choosing the plurality of quantization step values consists in choosing the same value for all the quantization steps.

  This optimizes the compression-distortion compromise while avoiding block effects.

  For the same purpose as that indicated above, the present invention also proposes a coding device for an image divided into a plurality of data blocks, each block undergoing a frequency transformation, this device being remarkable in that it comprises a module enabling a user to enter a plurality of quantization step values; a module for comparing, for each frequency of the current block, the current quantization step with a predetermined threshold; and a module replacing the step by the threshold, if the current quantization step exceeds the threshold, the replacement module providing a definitive quantization step.

  The present invention also provides a communication apparatus comprising a coding device as above.

  The present invention also provides a computer-readable information storage means or a microprocessor retaining instructions of a computer program, allowing the implementation of a coding method as above.

  The present invention also aims at a means for storing removable information, partially or totally, readable by a computer or a microprocessor retaining instructions of a computer program, allowing the implementation of a coding method such as above. .

  The present invention is also directed to a computer program product which can be loaded into a programmable apparatus, comprising instruction sequences for implementing a coding method as above, when this program is loaded and executed by the apparatus programmable.

  Since the particular features and advantages of the coding device, the communication apparatus, the storage means and the computer program product are similar to those of the coding method, they are not repeated here.

  Other aspects and advantages of the invention will appear on reading the following detailed description of a particular embodiment given by way of non-limiting example. The description refers to the accompanying drawings in which: - Figure 1 illustrates the general context of the present invention; FIG. 2 is a flowchart illustrating the main steps of a coding method according to the present invention, in a particular embodiment; and FIG. 3 schematically represents a device adapted to implement the present invention, in a particular embodiment.

  Figure 1 presents a simulation of the invention. A digital image 1 is coded by an encoder 2 which comprises a coding device according to the invention. This encoder produces a compressed file that can be transmitted or stored (phase 3 on the drawing). The compressed file is read by a decoder 4, which produces a decompressed (i.e., decoded) image 5.

  The main coding steps which are not part of the invention proper are recalled.

  The starting image is divided into blocks, preferably 8x8 pixels in size. Alternatively, one can use other block sizes, such as 4x4, 10X10, or even use only one block whose size is that of the image. A number of subsequent steps are applied to all blocks.

  A discrete cosine transformation is then applied to the current block. For more details on this operation, reference may be made to the book by WB PENNEBAKER and JL MITCHELL entitled "JPEG: Stil / Image Compression Standard Data", Van Nostrand Reinhold, 1993. Alternatively, it is possible to apply other types of transformations, such as a wavelet transformation.

  Then the current block is transformed into a single-dimensional block, by applying a technique known per se called "zigzag scan". This technique, described in the book by M. RABBANI and P. W. JONES entitled "Digital Image Compression Techniques", SPIE Opt. Eng. Press, Bellingham, Washington, 1991, consists of traversing all the samples of the block diagonally, starting at the top left and ending at the bottom right.

  The initial block is in the form of a two-dimensional data block of 8x8 samples, these samples being derived from the discrete cosine transformation. The samples are ordered by considering the diagonals of samples, oriented from the bottom left to the top right, starting with the diagonal at the top left to the bottom right. Inside a diagonal, samples are taken from the bottom left to the top right, or in the opposite direction. Among these two possible senses, the choice is made according to the following criterion: the direction must be the direction opposite to that used to traverse the diagonal considered previously. Following this direction of travel, the samples encountered are stored in a one-dimensional vector of 64 samples.

  The invention is not limited to this type of route. One can also transform the current block by traversing it in the lexicographic order, that is, line by line, to generate the new one-dimensional block. However, in the preferred embodiment, the zigzag scan is used.

  The following operation consists in quantifying the current block, using a quantization table the elaboration of which will be described in detail below with reference to FIG. 2. All the samples of the current block will be divided by a quantization step. Q in the table, then their value will be rounded to the nearest integer. The quantized block is therefore a list of 64 signed integer coefficients.

  Finally, the quantized block is encoded by entropy coding, for example by Huffman coding or by arithmetic coding.

  As shown in FIG. 2, a first step E1 of the coding method according to the present invention consists, for a user, of entering, for example by means of a keyboard, initial quantization steps.

  In the preferred embodiment, the type of coding used comprises a division of the image into blocks of size 8 × 8 pixels, with a discrete cosine transform or DCT (in English "Discrete Cosine Transform") 8 × 8 such as that used in FIG. JPEG standard.

  Thus, the final quantization table will have 64 quantization steps corresponding respectively to the 64 frequencies present in a DCT block. As a result, the initial quantization steps are also 64 in number.

  To enter these quantization steps, the user may have to consult a dialog window in which "no quantization" fields are proposed to him. II enters a value for each quantization step, in each field, then closes the dialog window.

  In the preferred embodiment, all initial quantization steps are considered to be equal. The user therefore enters a single value, for example 10, and this value will be considered as the initial quantization step of each of the 64 frequencies of the current block.

  As shown in block E4 in the drawing, an array of predetermined thresholds for the quantization steps is accessible by the image compression software.

  In the context of the invention, the table of predetermined thresholds generally comprises a number of values less than 64, that is to say that certain frequencies are assigned a threshold and others are not. In the preferred embodiment, the table has the following values: {9,10,10,11,11,11,00,00,00, ..., 00}. The symbol.0 means there is no threshold. Thus, in the preferred embodiment, the first frequency of the block includes a threshold of 9, the second a threshold of 10, the third a threshold of 10, the fourth a threshold of 11, the fifth a threshold of 11, the sixth a threshold of 11, and no threshold is associated with the frequencies of the seventh to the sixty-fourth.

  As a variant, the threshold table can be in the form {9,10,10,11,11,11}, that is to say that a threshold is assigned only to the 6 lowest frequencies of the current block. . Thus, one can initially choose a frequency of the block below which all the frequencies will have an associated threshold, and beyond which no threshold is associated with the frequencies.

  At the end of the input step E1 of the initial quantization steps, a step E2 consists in selecting the first frequency, which becomes the current frequency.

  Then during an E3 test, the value of the initial quantization step associated with the current frequency is compared with the threshold associated with the current frequency. If the value of the initial quantization step associated with the current frequency is less than or equal to the threshold associated with the current frequency (negative test E3), the following step E6 consists in storing the value of the current step in a table called the quantification. Otherwise, the next step E5 consists in storing the value of the current threshold in the quantization table. If no threshold is associated with the current frequency, the test E3 is followed by step E6.

  Steps E5 and E6 are followed by an E7 test, in which it is tested whether the current frequency is the last frequency, that is to say the 64th in the preferred embodiment. If this is the case (positive test E7), the next step is a step E8 encoding the image. Otherwise, the next step is a step E9 of selecting the next frequency.

  In step E9, the frequency that follows the current frequency becomes the new current frequency. This step is followed by the E3 test described above.

  As for step E8, it consists of coding the image itself, using the final quantization table previously constructed. This coding is conventional and is described for example in the work of W. B. PENNEBAKER and J. L. MITCHELL mentioned above.

  As described above, in the context of the particular embodiment considered here, the following list of thresholds is available: {9,10,10,11,11, 11,00,00,00, ..., 00} and assume that the initial quantization step for each frequency is 10. As a result, the final quantization table is: {9,10,10,10, ..., 10}. By way of example, if the initial quantization step for each frequency had been 15, the final table would have been: {9,10,10,11,11,11,15,15,15, ..., 15 }.

  Thanks to the present invention, the lowest frequencies can not have a value greater than a certain threshold. Now the Applicant has found that these low frequencies are those that generate discontinuity effects between blocks when the quantification is too important. According to the invention, the quantization is limited to values such that no block effect is generated.

  A device 10 implementing the coding method of the invention is illustrated in FIG.

  This device may be for example a microcomputer 10 connected to different peripherals, for example a digital camera 107 (or a scanner, or any means of acquiring or storing an image) connected to a graphics card and providing information to be treated according to the invention.

  The device 10 comprises a communication interface 112 connected to a network 113 capable of transmitting digital information. The device 10 also comprises a storage means 108 such as for example a hard disk. It also includes a floppy disk drive 109. The diskette 110, like the hard disk 108, may contain software layout data of the invention as well as the code of the invention which, once read by the device 10, will be stored in the hard disk 108. Alternatively, the program allowing the device to implement the invention may be stored in ROM 102 (called ROM in the drawing). The same is true for coding methods. In the second variant, the program can be received to be stored in a manner identical to that previously described via the communication network 113.

  The device 10 is connected to a microphone 111 via an input / output card 106. The data to be processed according to the invention will in this case be an audio signal.

  This same device has a screen 104 for viewing the information to be processed or to interface with the user who can set certain modes of treatment, using the keyboard 114 or any other means (mouse for example).

  The CPU 100 (called the CPU in the drawing) executes the instructions relating to the implementation of the invention, instructions stored in the read-only memory 102 or in the other storage elements. When powering up, the programs and processing methods stored in one of the (non-volatile) memories, for example the ROM 102, are transferred into the RAM RAM 103, which then contains the executable code of the invention as well as the variables necessary for the implementation of the invention. Alternatively, the methods of treatment can be stored in different places. Indeed, it is possible to improve the invention by adding new methods transmitted by the communication network 113 or by means of floppy disks 110. Of course, the diskettes can be replaced by any information medium such as CD -ROM or memory card.

  The communication bus 101 allows communication between the various sub-elements of the microcomputer 10 or linked to it. The representation of the bus 101 is not limiting and in particular, the central unit 100 is able to communicate instructions to any sub-element of the microcomputer 10 directly or via another sub-element of the microcomputer 10 .

  The device described here is likely to contain all or part of the treatment described in the invention.

Claims (14)

  A method of encoding an image divided into a plurality of data blocks, each block undergoing a frequency transform, said method being characterized in that it comprises the steps of: selecting (El) a plurality of values of no quantification; then, for a plurality of frequencies of the current block, comparing (E3) the current quantization step with a predetermined threshold; and - if the current quantization step exceeds said threshold, replace (E5) said step by said threshold, so as to obtain a definitive quantization step.   2. Method according to claim 1, characterized in that said plurality of frequencies consists of a set of frequencies lower than a predetermined frequency.   3. Method according to claim 1 or 2, characterized in that it further comprises a step of storing the value of the definitive quantization step in a quantization table.   4. Method according to claim 1, 2 or 3, characterized in that the blocks are 8x8 pixels in size.   5. Method according to any one of claims 1 to 4, characterized in that the step of choosing said plurality of quantization step values consists in choosing the same value for all the quantization steps.   6. A coding device for an image divided into a plurality of data blocks, each block undergoing a frequency transformation, said device being characterized in that it comprises: means enabling a user to input a plurality of data blocks; quantization step values; means for comparing, for a plurality of frequencies of the current block, the current quantization step with a predetermined threshold; and means for replacing said step by said threshold, if the current quantization step exceeds said threshold, said replacement means providing a definitive quantization step.   7. Device according to claim 6, characterized in that said plurality of frequencies consists of a set of frequencies below a predetermined frequency.   8. Device according to claim 6 or 7, characterized in that it further comprises means for storing the value of the definitive quantization step in a quantization table.   9. Device according to claim 6, 7 or 8, characterized in that the blocks are 8x8 pixels in size.   10. Device according to any one of claims 6 to 9, characterized in that the input means of said plurality of quantization step values are adapted to accept the same value for all quantization steps.   11. Communication apparatus, characterized in that it comprises a coding device according to any one of claims 6 to 10.   12. Computer-readable information storage medium or microprocessor retaining instructions of a computer program, characterized in that it allows the implementation of a coding method according to any one of claims 1 to 5.   13. Removable or partly computer-readable information storage medium readable by a computer or a microprocessor retaining instructions of a computer program, characterized in that it allows the implementation of a coding method according to FIG. any of claims 1 to 5.   14. Computer program product that can be loaded into a programmable device, characterized in that it comprises instruction sequences for implementing a coding method according to any one of claims 1 to 5, when this program is loaded and executed by the programmable device.