JP2010276686A - Image control device - Google Patents

Abstract

An image control apparatus for encoding image data so as to reduce a data transfer amount of a memory while suppressing deterioration of image quality.
A format conversion unit (10) divides each pixel component of image data (2) in a first format in which gradation of each pixel component is expressed by a plurality of bits into a predetermined bit group, and sets the bit group to a bit depth. Based on this, the image data is converted into image data of the second format. The image encoding unit 20 encodes the format-converted image data with a plurality of encoding methods, selects the encoding method according to the image quality index of the encoded data, and outputs the corresponding encoded data To do. The memory input control unit 30 allocates a memory area for storing a plurality of pixel components constituting the image data of the second format according to the selected encoding method, and stores the plurality of pixel components together. Therefore, it is possible to reduce the amount of memory transfer when decoding while suppressing deterioration in image quality.
[Selection] Figure 1

Description

  The present invention relates to encoding and decoding of image data, and in particular, an image in which image data is encoded so that a pixel value at an arbitrary position in the image data can be taken out while suppressing deterioration in image quality, and the encoded data is decoded. The present invention relates to a control device.

  In recent years, information processing apparatuses have become highly functional and multifunctional, and functions required for image control apparatuses mounted thereon have also diversified. For example, one of the functions is an RGB format conversion function for realizing conversion between low-quality display and high-quality display. As a technology related to this, there is an invention disclosed in Patent Document 1 below.

  An object of the invention disclosed in Patent Document 1 is to provide a memory controller, a display controller, and a memory control method capable of improving the use efficiency of a memory and simplifying addressing. The memory controller includes a dividing unit and an address generator.

When the division unit receives pixel data in which the number of bits of the first color component is I1 bits, the number of bits of the second color component is I2 bits, and the number of bits of the third color component is I3 bits, The input pixel data includes a basic data portion (J1 + J2 + J3 = J1 bits for the first color component, J2 bits for the second color component, and J3 bits for the third color component). 2 ^M ), the extended data portion (K1 + K2 + K3 = 2 ^N ) in which the number of bits of the first color component is K1 bits, the number of bits of the second color component is K2 bits, and the number of bits of the third color component is K3 bits. ) And split.

  The address generator writes the basic data portion in the basic data storage area of the memory and generates an access address for writing the extension data portion in the extension data storage area.

  When there is a display request for low image quality, only the basic data portion is read and displayed. When there is a display request for high image quality, the basic data portion and the extended data portion are read and displayed. As a result, the amount of image data transferred in the case of low image quality display can be reduced.

  Further, by controlling the division method of the bit group of image data and the storage method of the bit group in the memory, the memory usage efficiency and the memory access addressing can be easily performed.

  On the other hand, the following non-patent document 1 defines a JPEG (Joint Photographic Experts Group) encoding method and decoding method. In JPEG, when image data is encoded, the image is divided into blocks of 8 × 8 pixels, and discrete cosine transform is performed for each block to convert the image data from the spatial domain to the frequency domain. Then, the amount of information is reduced by quantizing the transformed data, and so-called variable length coding is performed by assigning codes having different lengths according to the level of occurrence probability of the data by entropy coding using Huffman codes. . When decoding the encoded image data, the reverse procedure of the encoding of the image data is executed.

  Non-Patent Document 2 below defines a standard for standard television broadcasting, corresponds to NTSC (National Television System Committee), PAL (Phase Alternation by Line), etc., and has a screen aspect ratio of 4: 3 and It corresponds to 16: 9. As color components, RGB 4: 4: 4 and color difference (Y, Cb, Cr) 4: 2: 2 are defined.

JP 2006-184792 A

International standard ISO / IEC 10918: 1994 ITU-R BT. 601 standard

  In Patent Document 1 described above, image data in RGB format is divided into predetermined bit groups to generate a basic data portion and an extended data portion and store them in a memory. However, since all the image data subjected to format conversion is stored in the memory, there is a problem that the memory capacity becomes enormous when handling high-resolution image data.

  In addition, when image data is encoded using JPEG defined in Non-Patent Document 1, the processing time is long because the processing is complicated. Further, since variable length coding is used, image data at an arbitrary position and encoded data do not correspond one-to-one. Therefore, in order to obtain image data at an arbitrary position, it is necessary to decode the encoded data in a certain range to obtain the image data, which causes problems such as complicated processing and long processing time. was there.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide an image control apparatus that encodes image data so as to reduce the data transfer amount of a memory while suppressing deterioration in image quality. That is.

  According to one embodiment of the present invention, an image control apparatus for encoding image data and storing it in a memory is provided. The format conversion unit divides each pixel component of the image data of the first format in which the gradation of each pixel component is expressed by a plurality of bits into predetermined bit groups, and combines the bit groups based on the bit depth. The image data is converted into the image data of the second format. The image encoding unit encodes the image data converted into the second format by the format conversion unit using a plurality of encoding methods, and selects the encoding method according to the image quality index of the encoded data. Output the corresponding encoded data. The memory input control unit allocates a memory area for storing a plurality of pixel components constituting the image data of the second format in accordance with the encoding method selected by the image encoding unit, and collects each of the plurality of pixel components. And stored in the encoded data memory.

  According to an embodiment of the present invention, the image encoding unit encodes the image data converted into the second format by the format conversion unit using a plurality of encoding methods, and the image quality index of the encoded data Since the encoding method is selected according to the above, it is possible to suppress the deterioration of the image quality. The memory input control unit allocates a memory area for storing a plurality of pixel components constituting the image data of the second format according to the encoding method selected by the image encoding unit, and Therefore, the amount of memory transfer when decoding is reduced.

1 is a block diagram illustrating a schematic configuration of an image control apparatus according to a first embodiment of the present invention. 3 is a diagram for explaining format conversion by a format conversion unit 10; FIG. 3 is a diagram illustrating a configuration example of an image encoding unit 20. FIG. It is a figure for demonstrating the kind of encoding system performed by the encoding part. It is a figure for demonstrating the thinning-out method shown in FIG.

(First embodiment)
FIG. 1 is a block diagram showing a schematic configuration of the image control apparatus according to the first embodiment of the present invention. The image control apparatus 1 encodes original image data 2 input from the outside, decodes the encoded data, and outputs the decoded image data 3 to the outside. The encoding unit 20, the memory input control unit 30, the encoded data memory 40, the encoding method information memory 50, the memory output control unit 60, and the image decoding unit 70 are included.

  FIG. 2 is a diagram for explaining format conversion by the format conversion unit 10. FIG. 2A shows the original image data 2 and, as an example, the case of the RGB888 format. In FIG. 2A, the RGB components are divided into nine bit groups “a” to “i”.

  The R component is divided into upper 3 bits “a”, intermediate 2 bits “b”, and lower 3 bits “c”. The G component is divided into upper 3 bits “d”, intermediate 3 bits “e”, and lower 2 bits “f”. The B component is divided into upper 2 bits “g”, intermediate 3 bits “h”, and lower 3 bits “i”.

  FIG. 2B shows an example of image data after format conversion, which is generated by combining divided bit groups based on the bit depth. The U component is generated by combining the upper bits of the RGB components, and is composed of “a”, “d”, and “g”. The C component is generated by combining the intermediate bits of the RGB components, and is composed of “b”, “e”, and “h”. The L component is generated by combining the lower bits of the RGB component, and is composed of “c”, “f”, and “i”.

  As shown in FIG. 2B, by converting a bitmap image in the RGB888 format to the UCL888 format, the RGB332 format can be expressed only by the U component. In addition, RGB565 format, RGB333 format, RGB444 format, etc. can be expressed by U component and C component. Of course, the RGB888 format can also be expressed by all the UCL components. Note that the number of bit group divisions, the combination method, the number of components after combination, and the like are not limited to those shown in FIG. 2, and an arbitrary number or combination can be selected.

  The image encoding unit 20 encodes the UCL format image data generated by the format conversion unit 10 at an arbitrary compression rate.

  FIG. 3 is a diagram illustrating a configuration example of the image encoding unit 20. The image encoding unit 20 receives the image data 22 output from the compression rate 21 and the format conversion unit 10, performs encoding, and outputs encoding method information 23 and encoded data 24. Parts A to D (200 to 203), PSNR calculation parts A to D (210 to 213), frame buffers A to D (220 to 223), a comparator 230, and a multiplexer 240.

  The encoding units A to D (200 to 203) receive the compression rate 21 and the image data 22, and perform encoding of the image data 22 by different methods.

  PSNR calculation units A to D (210 to 213) are provided corresponding to the encoding units A to D (200 to 203), respectively, and are output from the encoding units A to D (200 to 203). An image quality index PSNR (Peak Signal to Noise Rate) of the encoded data is calculated and output to the comparator 230, and the encoded data is stored in the frame buffers A to D (220 to 223), respectively. It should be noted that a method other than PSNR such as SSIM (Structural Similarity) can be used as an image quality index for determining image quality.

  The comparator 230 determines the image quality of each encoded data by comparing the PSNR values output from the PSNR calculation units A to D (210 to 213), and 2 bits of information on the encoding method having the best PSNR value. Is output to the multiplexer 240 and the encoding scheme information memory 50 as the encoding scheme information 23 of the encoding scheme.

  The multiplexer 240 receives the encoding method information 23 output from the comparator 230, and corresponds to the encoding method information 23 from the four encoded data stored in the frame buffers A to D (220 to 223). The encoded data is selected and output as encoded data 24.

  In the image encoding unit 20 shown in FIG. 3, encoding processing corresponding to one encoding method is performed by one set of encoding unit, PSNR calculation unit, and frame buffer, and four types are provided by parallelizing them. The encoding process according to the encoding method is performed in parallel. Note that the parallel number can be changed by increasing or decreasing the number of sets of the encoding unit, the PSNR calculation unit, and the frame buffer. Further, a process corresponding to a plurality of encoding methods may be pipelined by a set of encoding units, a PSNR calculation unit, and a frame buffer.

  FIG. 4 is a diagram for explaining the types of encoding methods performed by the encoding unit. FIG. 5 is a diagram for explaining the thinning method shown in FIG. FIG. 4 shows a case where a compression rate of 2/3 is designated as an example, and shows a case where image data is compressed by thinning. In FIG. 4, “4” is described when thinning out is not performed for each component of UCL, and “2” is described when thinning out 1/2.

  The encoding methods (1) to (3) shown in FIG. 4 are thinning methods in which the value of one pixel among a plurality of pixels is used as a representative and the value of this pixel is used as all values of the plurality of pixels. Image data is compressed to 2/3 as a whole by performing a representative decimation of 1/2 on two components in UCL.

  FIG. 5A shows an example in which representative thinning is performed, and “α” is selected as a representative from the values “α”, “β”, “γ”, and “δ” of four pixels. The case where “α” is used as all the values of the four pixels is shown.

  The encoding methods (4) to (6) shown in FIG. 4 are thinning methods in which an average value of a plurality of pixel values is calculated for each component and this average value is used as a value for each component of the plurality of pixels. Image data is compressed to 2/3 as a whole by performing 1/2 average decimation on two components in UCL.

  FIG. 5B shows an example of the case where average thinning is performed in units of components, and the average value “ε” of the values “α”, “β”, “γ”, “δ” of four pixels is calculated. In this example, “ε” is used as the value of all four pixels.

  The encoding methods (7) to (9) shown in FIG. 4 calculate the average value of the values of a plurality of pixels in units of bit groups, and combine these average values to use as values for each component of the plurality of pixels. It is. Image data is compressed to 2/3 as a whole by performing 1/2 average decimation on two components in UCL.

  FIG. 5C shows an example in the case where average thinning is performed on a bit group basis, and shows average thinning for the L component. An average value is calculated in units of bit groups of L components of four pixels “α”, “β”, “γ”, and “δ”.

  For example, when the average value of the bit group corresponding to “c” of the L component shown in FIG. 2B is calculated, “α_c”, “β_c”, “γ_c”, “δ_c” of four pixels are calculated. An average value “ζ_c” is calculated. Similarly, an average value “ζ_f” of the bit group corresponding to “f” of the L component and an average value ζ_i of the bit group corresponding to “i” of the L component are calculated. “Ζ_f” and ζ_i ”are combined to generate an average value“ ζ ”.

  The memory input control unit 30 determines the encoded data amount of each UCL component with reference to the encoding scheme information 23 output from the image encoding unit 20, and the memory area is free according to the encoded data amount. The encoded data 24 is efficiently stored by allocating the memory area so as not to be generated, and generating and outputting an address for the encoded data memory 40.

  For example, when the encoding method (1) shown in FIG. 4 is selected, the U component is not thinned out, and the C component and L component are thinned out to ½. At this time, the memory input control unit 30 assigns the memory area for storing the C component and the memory area for storing the L component as ½ to the memory area for storing the U component, respectively. Addressing is performed so that the L component is stored together in each memory area, and the L component is stored in the encoded data memory 40.

  The memory output control unit 60 controls the output of the encoded data stored in the encoded data memory 40 with reference to the encoding method information 23 and the display format information 80 stored in the encoding method information memory 50. To do.

  For example, if the display format information 80 indicates the RGB332 format, the memory output control unit 60 generates an address so as to read only the U component from the encoded data memory 40 based on the encoding method information 23, Output to the digitized data memory 40.

  If the display format information 80 indicates the RGB565 format, the memory output control unit 60 generates an address so as to read the U component and the C component from the encoded data memory 40 based on the encoding method information 23. And output to the encoded data memory 40.

  In this way, the components of the UCL format are collectively stored in the encoded data memory 40, and the memory output control unit 60 reads out only the encoded data necessary for the display format from the encoded data memory 40. The data transfer amount of the encoded data memory 40 is reduced.

  The image decoding unit 70 refers to the encoding method information 23 and the display format information 80 stored in the encoding method information memory 50 and decodes the encoded data output from the encoded data memory 40. The decoded image data 3 is output to the outside. For example, when image data is encoded by the encoding method (1) shown in FIG. 4, the image data is decoded by giving representative values to the thinned C component and L component.

  As described above, according to the image control apparatus of the present embodiment, the format conversion unit 10 generates image data in a format in which the gradation of each pixel component is expressed by a plurality of bits, such as the RBG format. Divide into bit groups. Since the UCL components are generated by combining the bit groups according to the bit depth, and the components are collectively stored in the memory, the image data in the display format such as the RGB332 format or the RGB565 format is read out. In addition, the amount of memory transfer can be reduced.

  In addition, since the image encoding unit 20 performs encoding corresponding to a plurality of encoding methods on the format-converted image data, the encoded data determined to have the best image quality is selected. It has become possible to prevent deterioration in image quality while encoding with the same encoded data amount.

  Further, since the image encoding unit 20 performs encoding by performing decimation at the same decimation rate on the entire image data in accordance with the compression rate 21, the arbitrary position of the image data, the encoded data, Is one-to-one, and image data at an arbitrary position can be easily acquired. Therefore, it can be used in a process in which reading / writing of image data at an arbitrary point frequently occurs, such as a BitBLT (Bit Block Transfer) process often used as a process of an embedded display device.

  Further, the memory input control unit 30 determines the encoded data amount of each UCL component according to the encoding method information 23 and allocates a memory area according to the encoded data amount. It became possible to store in the memory efficiently.

  In addition, since the memory output control unit 60 reads only necessary encoded data in accordance with the encoding method information 23 and the display format information 80, it is possible to reduce the data transfer amount of the memory.

  Further, since the image decoding unit 70 performs decoding in accordance with the encoding method information 23 and the display format information 80, a complicated decoding process such as JPEG is not necessary, and the processing time is shortened. It became possible.

(Second Embodiment)
In the first embodiment, the format conversion unit 10, the image encoding unit 20, the memory input control unit 30, the memory output control unit 60, and the image decoding unit 70 constituting the image control apparatus are realized by hardware. Explained.

  In the second embodiment of the present invention, some of the components of the image control apparatus are replaced by software. For example, consider a case where encoded data encoded by the image control apparatus according to the present embodiment is used as a source image of an embedded display device.

  Source image data used in an embedded display device may be generated in advance and stored in a memory, and real-time performance is often not required for image encoding processing. Therefore, in the present embodiment, among the configuration of the image control apparatus shown in FIG. 1, the format conversion unit 10, the image encoding unit 20, and the memory input control unit 30 are realized by software, and the other components are dedicated. Implemented with hardware. The components realized by software are realized by a CPU (Central Processing Unit) (not shown) executing a program stored in a memory.

  Since there is no time restriction in the image encoding by the image encoding unit 20, the types of encoding methods can be increased, or a method such as SIMM having a high processing amount but high accuracy can be used as an image quality index. Therefore, after image data is encoded in advance with a larger number of encoding methods, the image quality can be evaluated using a high-accuracy image quality index, and source image data with higher image quality and less data amount can be generated.

  As described above, according to the image control apparatus of the present embodiment, the format conversion unit 10, the image encoding unit 20, and the memory input control unit 30 are realized by software, and other components are implemented by hardware. I tried to do it. Therefore, in addition to the effects described in the first embodiment, the decoding process in the embedded display device can be performed in real time, and the memory for the source image data of the embedded display device can be reduced. It became.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

  DESCRIPTION OF SYMBOLS 1 Image control apparatus, 2 original image data, 3 Decoded image data, 10 Format conversion part, 20 Image encoding part, 21 Compression rate, 22 Image data, 23 Encoding system information, 24 Encoding data, 30 Memory input control 40, memory for encoded data, 50 memory for encoding system, 60 memory output control unit, 70 image decoding unit, 80 display format information, 200 to 203 encoding unit, 210 to 213 PSNR calculation unit, 220 to 223 Frame buffer, 230 comparator, 240 multiplexer.

Claims (7)

Each pixel component of the image data of the first format in which the gradation of each pixel component is expressed by a plurality of bits is divided into predetermined bit groups, and the bit groups are combined on the basis of the bit depth, and the second format Format conversion means for converting to image data;
The image data converted into the second format by the format conversion means is encoded by a plurality of encoding methods, and the corresponding encoded data is selected by selecting the encoding method according to the image quality index of the encoded data. Image encoding means for outputting
An image control apparatus comprising: memory input control means for controlling input of encoded data output by the image encoding means to a memory. The image data of the first format is composed of three pixel components,
The image data of the second format includes the three pixel components generated by dividing the three pixel components into three bit groups of upper bits, intermediate bits, and lower bits, respectively, and combining the corresponding bit groups. The image control device according to claim 1 configured. The image encoding means includes a plurality of encoding means for encoding with a plurality of encoding methods based on the image data converted into the second format by the format conversion means and the compression rate;
A plurality of image quality index calculating means provided corresponding to each of the plurality of encoding means and calculating an image quality index of the encoded data;
Comparing means for comparing the image quality index calculated by the plurality of image quality index calculating means and selecting an encoding method having the best image quality from the plurality of encoding methods;
The image control apparatus according to claim 1, further comprising a selection unit that selects and outputs encoded data corresponding to the encoding method selected by the comparison unit.   The memory input control means allocates a memory area for storing a plurality of pixel components constituting the image data of the second format according to the encoding method selected by the image encoding means, and the plurality of pixels The image control apparatus according to claim 1, wherein each component is stored together in the memory. The image control apparatus further encodes pixel components required for decoding from among the encoded data stored in the memory based on the encoding method and display format information selected by the image encoding means. Memory output control means for controlling the memory to output data;
The image control apparatus according to claim 1, further comprising: an image decoding unit that decodes encoded data output from the memory under the control of the memory output control unit.   The memory output control means includes, among encoded data stored in the memory according to the display format information, pixel components of only a bit group corresponding to upper bits, bit groups corresponding to upper bits and intermediate bits. The image control apparatus according to claim 5, wherein the memory is controlled to output either a pixel component or a pixel component of all bit groups.   The image control according to claim 5 or 6, wherein the image decoding means decodes the encoded data output from the memory based on the encoding method and display format information selected by the image encoding means. apparatus.

Images