JP2011109573A - Image compression device, and electronic camera - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve a conventional problem where a quantization parameter cannot be accurately obtained. <P>SOLUTION: The image compression device includes: an image data input means to input compression object image data; a first calculation means, having a quantization-like variation in advance where image data generally linearly varies in the relation with a file size, for obtaining a file size corresponding to the compression object image data for two predetermined quantization-like variations; a file size designation means to designate a target file size; a second calculation means to obtain a target quantization-like variation corresponding to the target file size in relation to the compression object image data based on the relation between the file size obtained by the first calculation means and the quantization-like variation; a storage means to store in advance a relation between a quantization-like variation and a quantization step; and an image compression means to obtain a quantization step corresponding to the target file size based on the target quantization-like variation obtained by the second calculation means from the relation between the quantization-like variation and the quantization step stored in the storage means, and compress the compression object image data using the obtained quantization step. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an image compression apparatus and an electronic camera.

  In a general electronic camera, a captured image is compressed and stored in a recording medium. In the image compression processing, for example, after orthogonally transformed coefficients are quantized with a predetermined quantization parameter, encoding for reducing redundant components is performed. However, since the distribution of spatial frequency components differs depending on the subject, the compressed image file capacity varies greatly even when the same quantization parameter is used. For this reason, it is difficult to predict the number of images that can be taken with the remaining capacity of the recording medium, and there is a problem that the usability as an electronic camera is not good. In view of this, a fixed-length compression technique that makes the compressed image file capacity substantially constant regardless of the type of subject has been studied (see, for example, Patent Document 1).

US Pat. No. 5,594,554

  However, since the compression rate and the quantization parameter do not necessarily have a linear relationship, the conventional technology has a problem that the quantization parameter for the target compression rate cannot be obtained with high accuracy from the test compression result.

  An object of the present invention is to provide an image compression apparatus and an electronic camera capable of accurately obtaining a quantization parameter for making a compressed image file capacity substantially constant.

  An image compression apparatus according to the present invention has in advance a quantized change amount in which image data changes almost linearly in relation to a file size and an image data input means for inputting image data to be compressed. First calculation means for obtaining a file size corresponding to the compression target image data for one quantized change amount, file size designation means for designating a target file size, and file size obtained by the first calculation means; A second computing means for obtaining a target quantizing change amount corresponding to the target file size with respect to the compression target image data based on a relationship with the quantizing change amount; and the quantizing change amount and the quantization step; The target calculated by the second calculation means from the relationship between the storage means for storing the relationship in advance and the quantization change amount stored in the storage means and the quantization step An image compression unit that obtains a quantization step corresponding to the target file size based on a child change amount, and compresses the image data to be compressed using the obtained quantization step; .

  In particular, the predetermined two quantization change amounts are set according to the target file size.

  Further, the first calculation means obtains a file size for one quantized change amount set according to the target file size, and another one quantized change determined according to the obtained file size. The file size with respect to the quantity is obtained.

  An image compression apparatus according to the present invention includes an image data input unit that inputs image data to be compressed, a storage unit that stores a correspondence relationship between a quantization index and a quantization change amount, and a specified quantization index. A compression unit that compresses the image data, a file size designation unit that designates a target file size, and the compression unit that uses the first quantization index and the second quantization index, Using the test compression means for obtaining the first file size and the second file size when compressed, respectively, and the correspondence relationship between the quantization index and the quantization change amount, the first quantization index is obtained. Quantization change amount for obtaining a corresponding first quantization change amount and a second quantization change amount corresponding to the second quantization index Output means, and the relationship between the quantization change amount of the image data and the file size, the first quantization change amount, the first file size, and the second quantization change amount. And a target quantization change amount calculating means for obtaining a target quantization change amount at the target file size according to the obtained relationship, and the quantization index and the quantization change amount. A target quantization index calculating means for obtaining a target quantization index from the target quantization change amount, and compressed image data obtained by compressing the image data by the compression means with the target quantization index. Image data output means for outputting.

  In particular, the first quantization index and the second quantization index are set according to the target file size.

  In addition, the test compression unit obtains a first file size by performing a first compression by the compression unit using a first quantization index corresponding to the target file size, and the first file size Using the second quantization index determined according to the above, the second compression is performed by the compression unit to obtain the second file size.

Furthermore, the relationship between the first quantized change amount and the first file size and the logarithmic value of the second quantized change amount and the second file size are substantially straight lines. To do.
In addition, the quantization change amount calculation unit may calculate a logarithmic value of each first quantization change amount and a second quantization change from the first quantization index and the second quantization index. A logarithmic value of the quantity, and the target quantized change calculating means calculates the logarithmic value of the first quantizing change quantity and the logarithmic value of the second quantizing change quantity and the first file. A logarithmic value of the target quantization change amount at the target file size is obtained from the relationship between the size and the second file size, and the target quantization index calculation means calculates the logarithmic value of the target quantization change amount. A target quantization index is obtained.

  An image compression apparatus according to the present invention includes an image data input unit that inputs image data to be compressed, a storage unit that stores a correspondence relationship between a quantization index and a quantization change amount, and a specified quantization index. A compression unit that compresses the image data; a target file size setting unit that sets a target file size of the compression unit; and a first unit when the image data is compressed with a first quantization index by the compression unit. Quantizing to obtain a first quantized change amount corresponding to the first quantized index using a test compression means for obtaining a file size of the image and a correspondence relationship between the quantized index and the quantized change amount A change amount calculating unit; and a relationship between the quantized change amount of the image data and the file size, the first quantized change amount and the first The target quantized change amount in the target file size is obtained from the file size and the initial quantized change amount set in advance and the initial file size corresponding to the initial quantized change amount, and according to the obtained relationship. Target quantization change amount calculation means to be obtained; target quantization index calculation means for obtaining a target quantization index from the target quantization change amount using a correspondence relationship between the quantization index and the quantization change amount; And image data output means for outputting compressed image data obtained by compressing the image data by the compression means at the target quantization index.

  In particular, the first quantization index is set according to the target file size.

  The first quantizing change amount and the first file size, and the relationship between the logarithmic value of the initial quantizing change amount and the logarithmic value of the initial file size are substantially straight lines. .

  Further, the quantization change amount calculation means obtains a logarithmic value of the first quantization change amount from the first quantization index, and the target quantization change amount calculation means calculates the first quantum change amount. The target from the relationship between the logarithmic value of the categorical change amount and the logarithm value of the preset initial quantizing change amount and the initial file size corresponding to the first file size and the logarithmic value of the initial quantizing change amount A logarithmic value of a target quantization change amount in a file size is obtained, and the target quantization index calculation means obtains a target quantization index from the logarithmic value of the target quantization change amount.

  An electronic camera according to the present invention includes the image compression device.

  In the present invention, a quantization parameter that makes the compressed image file capacity substantially constant can be obtained with high accuracy.

It is a block diagram which shows the structure of the electronic camera 101 which concerns on this embodiment. It is explanatory drawing which shows a response | compatibility with the quantization index in this embodiment, a quantization step, a pseudo quantization step, and its logarithmic value. 2 is a block diagram illustrating a configuration of an image processing unit 107. FIG. It is explanatory drawing which shows the relationship between a quantization index, a quantization step, a quantization step, and a compression rate. It is explanatory drawing which shows the relationship between the quantization index in this embodiment, a quantization step, and a quantization index and a pseudo quantization step. It is a flowchart which shows the image compression process which concerns on 1st Embodiment. It is explanatory drawing which shows how to obtain | require the quantization index and pseudo | simulation quantization step in this embodiment. It is explanatory drawing which shows the speed-up method of a test compression process. It is a flowchart which shows the image compression process of the modification of 2nd Embodiment. It is explanatory drawing which shows the principle of the image compression process which concerns on 3rd Embodiment. It is a flowchart which shows the image compression process which concerns on 3rd Embodiment.

  Embodiments relating to an image compression apparatus and an electronic camera according to the present invention will be described below. The electronic camera 101 described in the present embodiment is an example of an electronic camera equipped with the image compression apparatus according to the present invention.

[Configuration and Basic Operation of Electronic Camera 101]
First, the overall configuration and basic operation of the electronic camera 101 will be described. FIG. 1 is a block diagram illustrating a configuration of an electronic camera 101. The electronic camera 101 includes an optical system 102, a mechanical shutter 103, an image sensor 104, an A / D conversion unit 105, an image buffer 106, and an image processing unit. 107, a camera control unit 108, a memory 109, a display unit 110, a memory card IF (interface) 112, and an operation member 111. Here, the image processing unit 107 is a block including the image compression apparatus according to the present invention.

  In FIG. 1, the subject light incident on the optical system 102 is incident on the light receiving surface of the image sensor 104 via the mechanical shutter 103. Here, the optical system 102 includes a plurality of lenses such as a zoom lens and a focus lens, and includes a lens driving unit and a diaphragm controlled by the camera control unit 108.

  In the imaging element 104, pixels for performing photoelectric conversion on the light receiving surface are two-dimensionally arranged, and an analog image signal subjected to photoelectric conversion is output in units of pixels.

  The A / D conversion unit 105 converts an analog image signal output from the image sensor 104 into a digital value, and captures image data for one captured image into the image buffer 106. For example, when the resolution of the image sensor 104 is 1000 pixels × 1000 pixels, image data for 1 million pixels is taken into the image buffer 106. At this time, the image data taken into the image buffer 106 is called RAW data, and each pixel has one of RGB color components.

  The image buffer 106 is composed of a volatile high-speed memory, and not only temporarily stores the captured image output from the A / D conversion unit 105 but also used as a buffer when the image processing unit 107 performs image processing. Is done. Alternatively, it is also used as a display buffer when displaying a photographed image or a photographed image stored in the memory card 112 a connected to the memory card IF 112 on the display unit 110.

  The image processing unit 107 performs interpolation processing, white balance processing, gradation processing, color conversion processing, edge enhancement processing, and the like on the RAW data fetched into the image buffer 106. Furthermore, according to the parameter (quantization index) given from the camera control unit 108, for example, image compression processing conforming to the JPEGXR standard or the like is performed. The configuration of the image processing unit 107 and the quantization index will be described in detail later.

  The camera control unit 108 includes a CPU that operates according to a program stored therein, and controls the operation of the entire electronic camera 101. For example, the camera control unit 108 sets the shooting mode of the electronic camera 101 in response to a shooting mode selection dial (not shown) of the operation member 111 or a release button 111b, and controls the lens of the optical system 102 or the aperture. A subject image is picked up by the image pickup device 104 by controlling the control or the mechanical shutter 103. Then, the camera control unit 108 reads an image signal with a predetermined resolution from the image sensor 104, converts the image signal into a digital value by the A / D conversion unit 105, and captures image data for one screen into the image buffer 106. Furthermore, the camera control unit 108 instructs the image processing unit 107 to perform predetermined image processing on the image data captured in the image buffer 106. In particular, in the electronic camera 101 according to the present embodiment, the camera control unit 108 gives a parameter (quantization index) for performing compression processing to the image processing unit 107. At this time, the camera control unit 108 performs processing for obtaining a target quantization index corresponding to the target compression rate (target file size) in order to perform fixed length compression, and supplies the target quantization index to the image processing unit 107 to perform the main quantization index. Perform compression processing. Then, the image processing unit 107 adds a predetermined file name and header information to the image data after image compression, for example, JPEGXR data, and saves it in the memory card 112a via the memory card I / F 112, or shoots on the display unit 110. Display an image.

  The memory 109 is configured by a non-volatile semiconductor memory such as a flash memory, and stores parameters such as a shooting mode, exposure information, and focus information of the electronic camera 101, and the camera control unit 108 sets these parameters according to a user operation. Update as appropriate. In particular, in the present embodiment, in order to perform fixed-length compression processing, for example, a table indicating the correspondence between a quantization index, a quantization step, a pseudo-quantization step (quantization change amount) and the like as illustrated in FIG. 2 is manufactured. Pre-stored in the memory 109 at times. The quantization index, the quantization step, and the pseudo quantization step will be described in detail later.

  The display unit 110 is configured by a liquid crystal monitor or the like, and displays a captured image taken into the image buffer 106 by the camera control unit 108, a menu screen necessary for operating the electronic camera 101, and the like.

  In addition to the release button 111b, the operation member 111 includes a power button (not shown), a shooting mode selection dial (not shown), a white balance selection button (not shown), a cursor button (not shown), and the like. The user operates the electronic camera 101 using these operation buttons, and operation information by these operation buttons is output to the camera control unit 108. The camera control unit 108 controls the operation of the entire electronic camera 101 in accordance with operation information input from the operation member 111. In particular, in the present embodiment, the operation member 111 is provided with a target compression rate setting dial 111a for setting a target compression rate. The photographer uses the target compression ratio setting dial 111a to select one of the three types of FINE (1/4 compression ratio), NORMAL (1/8 compression ratio), and BASIC (1/16 compression ratio). Select. Note that setting the target compression rate corresponds to setting the target value of the data capacity (file size) of the compressed image data.

  The memory card IF 112 is an interface for connecting the memory card 112a to the electronic camera 101, and the camera control unit 108 writes image data to the memory card 112a via the memory card IF 112 or an image stored in the memory card 112a. Read data.

  The above is the configuration and basic operation of the electronic camera 101 common to the embodiments.

[Configuration and Operation of Image Processing Unit 107]
Next, the configuration and operation of the image processing unit 107 will be described in detail. FIG. 3 is a block diagram showing the configuration of the image processing unit 107 and the camera control unit 108 related to the image compression processing. In FIG. 3, the image processing unit 107 includes a white balance processing unit 201, a color interpolation processing unit 202, an edge enhancement processing unit 203, and a compression processing unit 204.

  The white balance processing unit 201 performs processing of multiplying the RGB component data of each pixel by a white balance coefficient so that the achromatic portion of the photographed subject is photographed as an achromatic image.

  The color interpolation processing unit 202 performs a process of interpolating RAW data in which each pixel has only one of RGB values and converting each pixel into RGB data having each RGB value.

  The edge enhancement processing unit 203 converts, for example, RGB data into luminance color difference data (Y, Cb, Cr), and performs edge enhancement processing on the luminance data Y.

  The compression processing unit 204 performs image compression processing on the luminance color difference data (Y, Cb, Cr) output from the edge enhancement processing unit 203 by a compression method according to a standard such as JPEG or JPEGXR. In particular, in this embodiment, it is assumed that image compression processing is performed by a compression method compliant with JPEGXR, and entropy coding is performed after quantization in a quantization step according to a specified quantization index. That is, in the present embodiment, the compression processing unit 204 can be regarded as a black box that compresses image data input with a designated quantization index (QPIndex) and outputs compressed image data (encoded data).

  Here, the configuration of the compression processing unit 204 will be described. The compression processing unit 204 includes an orthogonal transform unit 251, a quantization unit 252, and an entropy coding unit 253.

  The orthogonal transform unit 251 divides one input captured image into a plurality of sub-blocks, and performs orthogonal transform processing such as DCT (Discrete Cosine Transform) or wavelet transform for each sub-block to generate spatial frequency components. Convert and output coefficients for each frequency component. The orthogonal transformation process is performed for each of luminance (Y) and color difference (Cb, Cr), for example. In the following description of the compression processing, although it is not described in terms of sub-blocks and luminance / color difference dimensions, basically, a plurality of sub-blocks and luminance data / color difference data are subjected to the same processing to obtain one image. It is assumed that the entire image is compressed.

  The quantization unit 252 quantizes the coefficient output from the orthogonal transform unit 251 in a quantization step corresponding to the quantization index specified by the camera control unit 108. Note that in a normal electronic camera, the quantization step corresponding to the quantization index designated by the camera control unit 108 based on the correspondence between the quantization index (QPIndex) and the quantization step (QP) shown in FIG. The coefficient output from the orthogonal transform unit 251 is quantized. Here, in the quantization process, the compression rate changes according to the specified quantization index.For example, if the quantization index is increased, the quantization step also increases, so the compression rate increases, and conversely the quantization index decreases. As a result, the quantization step becomes smaller, and the compression rate becomes lower.

  The entropy encoding unit 253 performs entropy encoding that assigns a short code to a frequently occurring data sequence after performing difference processing of the quantized data output from the quantization unit 252 and code arrangement processing. The encoding process is also a process that affects the compression rate.

  As described above, the compression processing unit 204 includes the orthogonal transform unit 251, the quantization unit 252, and the entropy coding unit 253, and displays an image according to the quantization index (QPIndex) designated by the camera control unit 108. Perform compression processing.

  The image data after the image compression process (for example, JPEG image data) is stored as a captured image in the memory card 112a connected to the memory card IF 112 via the camera control unit 108.

  Here, the processing up to the white balance processing unit 201, the color interpolation processing unit 202, the edge enhancement processing unit 203, and the compression processing unit 204 is well-known image processing performed by a general electronic camera. A feature of the electronic camera 101 according to the present embodiment is that the camera control unit 108 is provided with a fixed length compression processing unit 205 for performing fixed length compression processing.

[Configuration and Operation of Fixed Length Compression Processing Unit 205]
Next, the configuration and operation of the fixed-length compression processing unit 205 will be described. First, the quantization index and the quantization step will be described. For example, the relationship between the quantization index (QPIndex) and the quantization step (QP) is as shown in the logarithmic graph of FIG. 4A, and the quantization step increases as the quantization index increases. The quantization step suddenly increases from a certain portion.

  Here, the quantization index is a parameter for performing image compression processing in accordance with the JPEGXR standard. For example, the relationship shown in FIG. 4A is determined in advance by the standard, and the quantization of the compression processing unit 204 is performed. The unit 252 quantizes the coefficient value orthogonally transformed in the quantization step corresponding to the quantization index designated by the camera control unit 108 in accordance with this standard.

  The quantization step means the width (quantization width) when the coefficient value is quantized, and the larger the quantization step, the coarser the coefficient value is decomposed, so the compression ratio increases but the image quality deteriorates. . Conversely, the smaller the quantization step, the finer the coefficient values are decomposed, so the compression ratio is lower but the image quality is improved. Note that the relationship between the compression rate and the quantization step differs depending on the type of image. This relationship is shown in FIG. FIG. 4B is a log-log graph showing the relationship of the compression rate (CR) when various types of images are compressed in the same quantization step (QP). The compression rate is the same even if the quantization step is the same depending on the type of image. Is different and has a distribution with a certain width. However, as shown in FIG. 4B, the overall tendency is that the compression rate decreases as the quantization step increases. In particular, in a region where the quantization step is small (for example, about 1 to 20), the relationship between the quantization step and the compression ratio is close to a straight line on the log-log graph, but in a region where the quantization step is large (for example, 50 or more). The relationship between the quantization step and the compression rate does not become a straight line on the log-log graph, and the inclination increases and the compression rate rapidly decreases.

  For this reason, for example, from the result of performing test compression by designating any two quantization indexes (corresponding to two quantization steps, respectively), the quantization step (corresponding to the target compression ratio using the linear interpolation method) ( And the corresponding quantization index) are difficult to estimate. Therefore, in the electronic camera 101 according to the present embodiment, the fixed length compression processing unit 205 performs a special quantization step (pseudo quantization step) in which the compression rate and the quantization step are proportional to each other in a log-log graph. By using, the quantization step (and corresponding quantization index) corresponding to the target compression rate can be estimated with high accuracy.

Next, the configuration and operation of the fixed length compression processing unit 205 will be described. The fixed-length compression processing unit 205 includes a target compression rate setting unit 261, a compression rate checking unit (capacity checking unit) 262, a pseudo quantization width calculation unit 263, a target pseudo quantization width calculation unit 264, and a target quantization. It is comprised with the index calculation part 265. Hereinafter, each part will be described in order.
[Target compression rate setting unit 261]
The target compression rate setting unit 261 sets the compression rate selected by the user using the target compression rate setting dial 111a of the operation member 111 as the target compression rate of the fixed length compression process. Although the target compression rate may be designated numerically, in the electronic camera 101 according to the present embodiment, for example, the target compression rate setting dial 111a is selected from three types of FINE, NORMAL, and BASIC. When the target compression ratio setting dial 111a is selected as FINE, the compression ratio is 1/4, when NORMAL is selected, the compression ratio is 1/8, and when BASIC is selected, the compression ratio is 1/16. As shown in FIG.
[Compression rate survey unit 262]
The compression rate checking unit 262 may be referred to as, for example, a capacity checking unit, and temporarily selects a quantization index for test compression, and gives this to the quantization unit 252 of the compression processing unit 204 to perform test compression. Then, the capacity of the image data after the test compression is checked to obtain the compression rate in the quantization index. Here, the compression rate includes the capacity of image data before compression processing input by the orthogonal transform unit 251 of the compression processing unit 204 in FIG. 3 and encoded data after compression processing (image compression) output by the entropy encoding unit 253. Data) capacity. For example, if the data capacity (file size) before compression is A (MB (megabytes)) and the data capacity (file size) after compression is B (MB), the compression rate (CR) is CR = B / A. Can be requested.

  Next, how to determine the quantization index for test compression will be described. First, a quantization step (QP) for test compression corresponding to the target compression rate (CR_T) set by the target compression rate setting unit 261 is obtained. This can be read from the graph showing the relationship between the quantization step (QP) and the compression ratio (CR) in FIG. In actual processing, for example, the relationship between the quantization step (QP) and the compression rate (CR) may be tabulated and stored in the memory 109. However, as described above, the quantization step (QP) is distributed with a certain width depending on the image type even with the same compression rate (CR), and therefore, a value having a statistically high probability is selected from the distribution range. . For example, in FIG. 4B, when the compression ratio is FINE (compression ratio ¼), the quantization step (QP) distribution range corresponding to the compression ratio ¼ (0.25) is about 6 to 60. The quantization step (QP) is selected from within this range. For example, when one quantization step is selected, the vicinity of 30 near the center of the distribution is selected, and when two quantization steps are selected, two points near 20 and 40 near the center of the distribution are selected. Similarly, in FIG. 4B, the range of quantization steps corresponding to NORMAL (compression ratio 1/8) is about 20 to 100, and the range of quantization steps corresponding to BASIC (compression ratio 1/16) is 50. Therefore, a quantization step (QP) having a statistically high probability is selected from within each distribution range. In the actual processing, as described above, a table indicating the distribution range of the quantization step (QP) for each target compression rate selectable by the target compression rate setting dial 111a or a statistically high probability value is used. It may be obtained at the time of manufacture or the like and stored in the memory 109, and this table may be referred to.

  Next, a test compression quantization index (QPIndex) corresponding to the selected test compression quantization step (QP) is obtained. This can be obtained from the relationship between the quantization step (QP) and the quantization index (QPIndex) shown in FIG. The quantization step (QP) and the quantization index (QPIndex) correspond one-to-one. For example, if the compression rate is FINE (compression rate ¼) and two quantization steps 22 and 42 are selected in the previous process, the quantization step (QP) is 22 in FIG. The quantization index (QPIndex) is between 16 and 32, and is 22 because Delta = 1. Similarly, when the quantization step (QP) is 42, the quantization index (QPIndex) is between 32 and 48, and since Delta = 2, it is 37. When the quantization index (QPIndex) does not become an integer, for example, the decimal part is rounded off.

In this way, the compression rate examining unit 262 can determine the test compression quantization index (QPIndex) corresponding to the designated target compression rate (CR_T), and uses this test compression quantization index. Then, test compression is performed to obtain a compression rate in the quantization index.
[Pseudo Quantization Width Calculation Unit 263]
The pseudo quantization width calculation unit 263 obtains a pseudo quantization step (QP2) corresponding to the test compression quantization index (QPIndex). Here, a pseudo-quantization step that replaces the quantization step (QP) so that the relationship between the quantization step (QP) and the compression ratio (CR) shown in FIG. 4B approaches a straight line on the log-log graph. (QP2) is calculated. The pseudo quantization step (QP2) is a quantization width in which the relationship with the compression rate (CR) is close to a straight line on the log-log graph, for example, as shown in FIG. The pseudo-quantization step (QP2) in FIG. 4 (c) is obtained so as to be exponentially larger in the region where the quantization width is about 30 or more than the quantization step (QP) in FIG. 4 (b). . 4C is a log-log graph showing an example of the result of actual measurement of the relationship between the pseudo-quantization step (QP2) and the compression rate (CR), similar to FIG. 4B. Even in the same pseudo-quantization step, the compression rate is different and the distribution has a certain width.

  Here, the relationship between the quantization step (QP) and the pseudo quantization step (QP2) will be described. The quantization unit 252 of the compression processing unit 204 uses the relationship between the quantization index (QPIndex) and the quantization step (QP) as shown in FIG. Quantization is performed in a quantization step (QP) corresponding to QPIndex). Here, FIG. 5A corresponds to FIG. In FIG. 5A, the quantization index (QPIndex) is divided into a plurality of sections increasing by 16, and the relationship with the corresponding quantization step (QP) is shown in each section. It is an approximate line graph.

  Specifically, in FIG. 5A, the change of the quantization step (QP) with respect to the quantization index (QPIndex) becomes QP = QPIndex in the interval of QPIndex ≦ 16, and every time QPIndex increases by 1 when 16 <QPIndex ≦ 32. QP also changes with increment (Delta) = 1. In other words, QP and QPIndex have a linear relationship in a section where QPIndex ≦ 32. Further, in the section of 32 <QPIndex ≦ 48, the value changes with increment (Delta) = 2, and whenever QPIndex increases by 1, QP increases by 2, so that QP when QPIndex is 48 becomes 64. Similarly, in the interval of 48 <QPIndex ≦ 64, the value changes with an increment (Delta) = 4. Since QP increases by 4 every time QPIndex increases by 1, QP becomes 128 when QPIndex is 64. Thus, the increment (Delta) corresponds to the slope, and thereafter, every time QPIndex increases by 16, the increment of QP increases by a power of 2.

Here, in FIG. 5A, when a variable i (i is an integer) indicates a QPIndex interval, for example, the relationship between the variable i and the QPIndex interval and the increment (Delta) is as follows.
i = 0: Delta = 1 substantially in a section where QPIndex is 1 or more and 16 or less
i = 1: Delta = 1 in the interval where QPIndex is greater than 16 and less than or equal to 32
i = 2: Delta = 2 in the interval where QPIndex is greater than 32 and less than or equal to 48
i = 3: In a section where QPIndex is greater than 48 and less than or equal to 64, Delta = 4
i = 4: Delta = 8 in a section where QPIndex is greater than 64 and less than or equal to 80
Hereinafter, it is similarly defined. If increment (Delta) is represented by variable i, it can be represented by Delta = 2 ⁽ⁱ⁻¹⁾ (when i ≧ 1).

  As described above, the quantization unit 252 of the compression processing unit 204 uses the relationship between the quantization index (QPIndex) and the quantization step (QP) in FIG. A quantization step corresponding to the index is obtained and quantization processing is performed. 5A corresponds to the QP in the table of FIG. 2 and is stored in advance in the memory 109 as a parameter.

  On the other hand, the pseudo quantization step (QP2) used by the fixed length compression processing unit 205, which is a feature of the present embodiment, to obtain the target quantization index (QPIndex_T) for the target compression ratio (CR_T) is shown in FIG. As shown, the quantization width is larger than the relationship between the quantization index (QPIndex) and the quantization step (QP) in FIG. In FIG. 5B, the quantization index (QPIndex) is divided into a plurality of sections that increase by 16 as in FIG. 5A, and the relationship with the corresponding pseudo quantization step (QP2) in each section is as follows. This is a line graph approximated by a straight line for each section. However, in FIG. 5B, since the increment of the quantization width is larger than that in FIG. 5A, the relationship between the compression ratio (CR) and the pseudo quantization step (QP2) is as shown in FIG. 4C. Almost straight.

  Specifically, in FIG. 5B, the change of the pseudo quantization step (QP2) with respect to the quantization index (QPIndex) is QP2 = QPIndex in the interval of QPIndex ≦ 16, and QPIndex is increased by 1 when 16 <QPIndex ≦ 32. Every time QP2 also changes with increment (Delta) = 1. That is, in the section of QPIndex ≦ 32, QP2 and QPIndex have a linear relationship as in FIG. 5A differs from FIG. 5A in a section where QPIndex is larger than 32. For example, in a section where 32 <QPIndex ≦ 48, it changes with increment (Delta) = 5, and as QPIndex increases by 1, QP2 increases by 5. QP2 when is 48 is 112. Similarly, in the section of 48 <QPIndex ≦ 64, the value changes with increment (Delta) = 25, and every time QPIndex increases by 1, QP2 increases by 25, so QP2 when 512 is QP2 becomes 512. Thus, as in FIG. 5A, the increment (Delta) corresponds to the slope, and thereafter, every time QPIndex increases by 16, the increment of QP2 increases by a power of 5.

Here, in FIG. 5B, when a variable i (i is an integer) indicates a QPIndex interval, for example, the relationship between the variable i and the QPIndex interval and the increment (Delta) is as follows.
i = 0: Delta = 1 substantially in a section where QPIndex is 1 or more and 16 or less
i = 1: Delta = 1 in the interval where QPIndex is greater than 16 and less than or equal to 32
i = 2: Delta = 5 in the interval where QPIndex is greater than 32 and less than or equal to 48
i = 3: In a section where QPIndex is greater than 48 and less than or equal to 64, Delta = 25
i = 4: Delta = 125 in a section where QPIndex is greater than 64 and less than or equal to 80
Hereinafter, it is similarly defined. Similarly to FIG. 5A, when increment (Delta) is represented by a variable i, it can be represented by Delta = 5 ⁽ⁱ⁻¹⁾ (when i ≧ 1).

  In the present embodiment, the quantization index (QPIndex) and the pseudo quantization step (QP2) such that the relationship between the compression ratio (CR) and the pseudo quantization step (QP2) is close to a straight line as shown in FIG. 4C. The target quantization index (QPIndex_T) with respect to the target compression rate (CR_T) can be obtained with high accuracy. 5B corresponds to QP2 in the table of FIG. 2, and is stored in advance in the memory 109 as a parameter. In the above description, the increment (Delta) is a power of 5. However, it is a value obtained experimentally by performing compression processing on images obtained by photographing various types of subjects, and the increment (Delta). Does not need to be a power of 5, and the quantization index (QPIndex) and the pseudo quantum such that the relationship between the compression ratio (CR) and the pseudo quantization step (QP2) is close to a straight line as shown in FIG. Any value can be used as long as the relationship of the quantization step (QP2) is obtained. Also, the increment (Delta) in which the relationship between the compression rate (CR) and the pseudo quantization step (QP2) is close to a straight line may differ depending on the model of the camera and the resolution of the image to be captured. An appropriate increment (Delta) for each resolution may be stored in the memory 109 in advance so that it can be automatically (or manually) selected according to the shooting mode.

In this way, the pseudo quantization width calculation unit 263 performs a process of obtaining a pseudo quantization step (QP2) corresponding to the test compression quantization index (QPIndex).
[Target pseudo-quantization width calculation unit 264]
The target pseudo quantization width calculation unit 264 includes a pseudo quantization step (QP2) corresponding to a test compression quantization index (QPIndex) and a compression obtained by test compression using the test compression quantization index (QPIndex). The target pseudo-quantization step (QP2_T) for obtaining the target compression rate (CR_T) is calculated using the rate (CR). Here, since the relationship between the compression ratio (CR) and the pseudo quantization step (QP2) can be regarded as a substantially straight line in the log-log graph as described in FIG. 4C, the linear interpolation method is applied. Thus, the target pseudo quantization step (QP2_T) for the target compression rate (CR_T) can be obtained with high accuracy. A specific processing flow will be described in detail later.
[Target quantization index calculation unit 265]
The target quantization index calculation unit 265 calculates a target quantization index (QPIndex_T) corresponding to the target compression rate (CR_T) from the target pseudo quantization step (QP2_T) obtained by the target pseudo quantization width calculation unit 264. Here, the target quantization index (QPIndex_T) is obtained using the graph of FIG. 5B showing the relationship between the quantization index (QPIndex) and the pseudo quantization step (QP2) or an equivalent table. A specific processing flow will be described in detail later.

As described above, the target compression rate setting unit 261, the compression rate investigation unit (capacity investigation unit) 262, the pseudo quantization width calculation unit 263, and the target pseudo quantization width calculation unit 264 constituting the fixed length compression processing unit 205. And the target quantization index calculation unit 265 operates.
(First embodiment)
Next, a first embodiment of the image compression apparatus and the electronic camera 101 according to the present invention will be described. The configuration of the electronic camera 101 is the same as the configuration described in FIG. In the present embodiment, a specific processing flow of the fixed-length compression processing unit 205 described above will be described.

  FIG. 6 is a flowchart of an image compression process centering on a process for obtaining a quantization index corresponding to the set target compression ratio, and performing a main compression process by obtaining a quantization index corresponding to the target compression ratio. Save to the memory card 112a.

Hereinafter, it demonstrates in order according to the flowchart of FIG.
(Step S101) As described above, the target compression rate setting unit 261 of the fixed length compression processing unit 205 presets the compression rate selected by the target compression rate setting dial 111a of the operation member 111 as the target compression rate. . Note that this step does not necessarily have to be performed for each shooting, and may be performed, for example, when the power is turned on or setting before shooting.
(Step S <b> 102) When the release button 111 b is pressed with the operation member 111 of the electronic camera 101, the image data photographed by the image sensor 104 is taken into the image buffer 106, the white balance processing unit 201 of the image processing unit 107, color After the image processing is performed by the interpolation processing unit 202 and the edge enhancement processing unit 203, the image data is stored in the image buffer 106 again. At this time, the image data input to the compression processing unit 204 is stored in the image buffer 106.
(Step S103) The compression rate examining unit 262 selects two test compression quantization indexes (QPIndex_1 and QPIndex_2) corresponding to the target compression rate (CR_T) with reference to a table stored in the memory 109.

  Then, the compression rate examining unit 262 gives the quantization index (QPIndex_1 and QPIndex_2) for test compression to the quantization unit 252 of the compression processing unit 204, and performs test compression on the image data input to the compression processing unit 204. Do. In the test compression, the coefficients orthogonally transformed by the orthogonal transformation unit 251 are quantized by two quantization indexes (QPIndex_1 and QPIndex_2), respectively, and the entropy coding unit 253 encodes each quantized data. Then, the compression rate checking unit 262 acquires the respective encoded data output from the entropy encoding unit 253 and checks the capacity of the compressed image data.

Here, for example, assuming that the volumes of compressed image data corresponding to two quantization indexes (QPIndex_1 and QPIndex_2) are VM1 and VM2, respectively, and the volume of image data input to the compression processing unit 204 is VM0, two quantizations Respective compression ratios (CR_1 and CR_2) corresponding to the indexes (QPIndex_1 and QPIndex_2) are obtained as (Expression 1) and (Expression 2).
CR_1 = VM1 / VM0 (Formula 1)
CR_2 = VM2 / VM0 (Formula 2)
(Step S104) The pseudo quantization width calculation unit 263 obtains pseudo quantization steps (QP2_1 and QP2_2) corresponding to the two quantization indexes (QPIndex_1 and QPIndex_2) for test compression. This method is obtained using a graph or table showing the relationship between the quantization index (QPIndex) and the pseudo quantization step (QP2) described in FIG. This is shown in FIG. FIG. 7 (a) is the same graph as FIG. 5 (b). Here, the function of FIG. 7A can be expressed as (Equation 3).
QP2 = g (QPIndex) (Formula 3)
In the example of FIG. 7A, the case where the quantization index QPIndex_1 = 36 and the quantization index QPIndex_2 = 58 is shown. Since the pseudo quantization step QP2_1 for QPIndex_1 = 36 is located in the section of Delta = 5 as described in FIG. 5B, QP2_1 = 32 + (36−32) · 5 = 52. Similarly, since the pseudo quantization step QP2_2 for QPIndex_1 = 58 is located in the section of Delta = 25, QP2_1 = 112 + (58−48) · 25 = 362.

Here, an example of program processing in C language for examining which section the two quantization indexes (QPIndex_1 and QPIndex_2) belong to in the above processing is shown below.
[Program (1)]
for (i = 0; i <16; i ++) {
if (QPIndex_x> = QPIndex [i] && QPIndex_x <= QPIndex [i + 1]) {
QP2_x = (QP2 [i + 1] -QP2 [i]) / (QPIndex [i + 1] -QPIndex [i]) *
(QPIndex_x-QPIndex [i]) + QP2 [i];
break;
}
}
Here, in the above program (1), the variable i is the same as the variable i described with reference to FIG. 5B and indicates a QPIndex section, and QPIndex_x corresponds to QPIndex_1 and QPIndex_2 to be checked. In this embodiment, the range of QPIndex is 0 to 255, and QPIndex is divided into 16 units, so the variable i indicating the interval is 1 to 16.

In this manner, the pseudo quantization width calculation unit 263 obtains pseudo quantization indexes (QP2_1 and QP2_2) corresponding to the two quantization indexes for test compression (QPIndex_1 and QPIndex_2).
(Step S105) The target pseudo-quantization width calculation unit 264 uses the pseudo-quantization steps (QP2_1 and QP2_2) obtained by the test compression and the compression ratios (CR_1 and CR_2) respectively corresponding to the target compression ratios (CR_T). The target pseudo-quantization step (QP2_T) is calculated. Here, since the relationship between the compression ratio (CR) and the pseudo quantization step (QP2) can be regarded as a straight line in the log-log graph as described in FIG. 4, the target compression is applied by applying the linear interpolation method. The target pseudo-quantization step (QP2_T) for the rate (CR_T) can be determined (Equation 4). Since the linear relationship is established on the log-log graph, first, a target pseudo-quantization step (log (QP2_T)) of logarithmic value is obtained, and QP2_T is calculated.

  For example, in (Expression 4), the target compression rate (CR_T) is CR_T = 0.25, the pseudo quantization steps (QP2_1 and QP2_2) are QP2_1 = 52 and QP2_2 = 362, and the compression rates (CR_1 and CR_2) are CR_1, respectively. = 0.10 and CR_2 = 0.58, the target pseudo-quantization step (QP2_T) is QP2_T≈132.

In this way, the pseudo quantization width calculation unit 263 uses the pseudo quantization steps (QP2_1 and QP2_2) obtained by the test compression and the compression rates (CR_1 and CR_2) corresponding to the respective steps to the target compression rate (CR_T). The target pseudo-quantization step (QP2_T) for is calculated.
(Step S106) The target quantization index calculation unit 265 calculates a target quantization index (QPIndex_T) corresponding to the target pseudo quantization step (QP2_T) obtained in step S105.

Here, the target quantization index (QPIndex_T) is obtained using the graph of FIG. 5B showing the relationship between the quantization index (QPIndex) and the pseudo quantization step (QP2) or an equivalent table. This state is shown in FIG. FIG. 7 (b) is the same graph as FIG. 5 (b). Moreover, the function of FIG.7 (b) can be represented like (Formula 5) as an inverse function of (Formula 3) demonstrated previously.
QPIndex_T = g ⁻¹ (QP2_T) (Formula 5)
The example of FIG. 7B shows a case where the target pseudo quantization step (QP2_T) obtained in step S105 is QP2_T≈132. The target quantization index QPIndex_T for QP2_T = 132 is QPIndex_T = (132−112) / 5 + 48 = 52.

Here, an example of program processing in C language for examining which section the target quantization index (QPIndex_T) belongs to in the above processing is shown below.
[Program (2)]
for (i = 0; i <16; i ++) {
if (QP2_T> = QP2 [i] && QP2_T <= QP2 [i + 1]) {
QPIndex_T = (QPIndex [i + 1] -QPIndex [i]) / (QP2 [i + 1] -QP2 [i]) *
(QP2_T-QP2 [i]) + QPIndex [i];
break;
}
}
Here, in the program (2), the variable i is the same as the variable i described with reference to FIG. 5B and indicates the QPIndex section, and QPIndex [i] is the QPIndex value of the start position of the section i = (16 * Corresponds to i). For example, since QP2_T = 132 corresponds to i = 3 (section 3), it is between QP2 [3] = 112 and QP2 [4] = 512, and the QPIndex value at the start position of this section = 16 * 3 It can be seen that = 48. That is, the target quantization index (QPIndex_T) is at i = 3 (section 3), and is a value between QPIndex = 48 and QPIndex = 64. If the interval in which the target quantization index (QPIndex_T) is located is known, the increment Delta in that interval can be known, so that the target quantization index (QPIndex_T) can be interpolated by proportional distribution.

In this way, the target quantization index calculation unit 265 calculates the target quantization index (QPIndex_T) corresponding to the target pseudo quantization step (QP2_T) obtained in step S105.
(Step S107) The compression processing unit 204 gives the target quantization index (QPIndex_T) obtained in step S106 to the quantization unit 252, and performs main compression on the image data input to the compression processing unit 204. In the main compression, the coefficient orthogonally transformed by the orthogonal transform unit 251 is quantized with the target quantization index (QPIndex_T), the quantized data is encoded by the entropy encoding unit 253, and is output as compressed image data.
(Step S108) The camera control unit 108 stores the compressed image data output from the compression processing unit 204 of the image processing unit 107 in the memory card 112a via the memory card I / F 112 as photographed image data.

  Actually, in the image compression processing, as shown in FIG. 8A, one photographed image is divided into a plurality of sub-blocks, and the above-described compression processing is performed for each block.

  In this way, the electronic camera 101 according to the first embodiment can accurately obtain the quantization index from the test compression result so that the set target compression rate is obtained. As a result, the volume of the compressed image data can be made substantially constant, so that the number of images that can be captured with the target compression rate selected according to the remaining capacity of the memory card 112a can be grasped. Usability of the camera 101 is improved.

  In the first embodiment, since test compression is performed using two quantization indexes (QPIndex_1 and QPIndex_2), it is necessary to perform test compression twice. However, as shown in FIG. When dividing a captured image into a plurality of sub-blocks and performing compression processing for each block, as shown in FIG. 8B, a block to which a quantization index (QPIndex_1) is applied, and a quantization index By performing the quantization separately for the blocks to which (QPIndex_2) is applied, it is possible to perform one test compression. In the example of FIG. 8B, since the block A that applies the quantization index (QPIndex_1) and the block B that applies the quantization index (QPIndex_2) are divided in a zigzag manner, the compression ratio examining unit 262 When obtaining the capacity of the compressed image data, it is calculated separately for the block A and the block B, and the compression rate in each block is calculated. By processing in this way, the processing amount and processing time can be reduced.

(Second Embodiment)
Next, a second embodiment of the image compression apparatus and the electronic camera 101 according to the present invention will be described. The configuration of the electronic camera 101 is the same as that described with reference to FIG. In the present embodiment, the processing of the fixed-length compression processing unit 205 described above is slightly different from that in the first embodiment.

FIG. 9 is a flowchart of the image compression process centering on the process for obtaining the quantization index corresponding to the set target compression rate in the second embodiment. The process of step S103 in FIG. After inputting the image data in S102, the following processing is executed instead of executing the processing of Step S103.
(Step S103a) The compression rate examining unit 262 first selects one test compression quantization index (QPIndex_1) corresponding to the target compression rate (CR_T) with reference to the table stored in the memory 109. As described above, the selection method selects the quantization index value having the highest probability.

Then, the compression rate examining unit 262 gives the quantization index (QPIndex_1) for test compression to the quantization unit 252 of the compression processing unit 204, and performs test compression on the image data input to the compression processing unit 204. In the test compression, the coefficients orthogonally transformed by the orthogonal transform unit 251 are quantized with a quantization index (QPIndex_1), and the entropy coding unit 253 codes each quantized data. The compression rate checking unit 262 acquires the encoded data output from the entropy encoding unit 253 and checks the capacity of the compressed image data. For example, when the capacity of the compressed image data corresponding to the quantization index (QPIndex_1) is VM1, and the capacity of the image data input to the compression processing unit 204 is VM0, it is expressed by (Expression 1) described in the first embodiment. The compression rate (CR_1) corresponding to the quantization index (QPIndex_1) can be obtained.
(Step S103b) The compression rate examining unit 262 selects the second quantization index (QPIndex_2) for test compression according to the compression rate (CR_1). For example, when the compression rate (CR_1) in the first quantization index for test compression (QPIndex_1) obtained in step S103a is smaller than the target compression rate (CR_T), the selection method is for the second test compression. As the quantization index (QPIndex_2), a value larger than QPIndex_1 is selected. Conversely, when the compression rate (CR_1) in the first test compression quantization index (QPIndex_1) is larger than the target compression rate (CR_T), the second test compression quantization index (QPIndex_2) is obtained from QPIndex_1. Choose a smaller value. As a result, a quantization index (QPIndex_2) for test compression that is as close as possible to the target compression ratio (CR_T) can be selected, but it is not always necessary to perform this processing, and the compression ratio (CR_1) obtained in the first test compression The second quantization index for test compression (QPIndex_2) may be selected by another method using the above.

  Then, the compression ratio examining unit 262 gives the second quantization index (QPIndex_2) for test compression to the quantization unit 252 of the compression processing unit 204, and performs test compression on the image data input to the compression processing unit 204. As in step S103a, the compression ratio (CR_2) corresponding to the quantization index (QPIndex_2) is obtained using (Expression 2) described in the first embodiment.

  In FIG. 9, the processing after step S104 is performed in the same manner as in FIG. 6 described in the first embodiment.

  As described above, the electronic camera 101 according to the second embodiment performs the test compression in two steps, thereby obtaining the two quantization indexes (QPIndex_1 and QPIndex_2) to be selected for the test compression as the first. Since a more appropriate value can be selected than in the embodiment, the target quantization index (QPIndex_T) for performing the main compression processing so as to be close to the set target compression rate can be obtained with higher accuracy. As a result, the volume of the compressed image data can be made substantially constant, so that the number of images that can be captured with the target compression rate selected according to the remaining capacity of the memory card 112a can be grasped. Usability of the camera 101 is improved.

(Third embodiment)
Next, a third embodiment of the image compression apparatus and the electronic camera 101 according to the present invention will be described. The configuration of the electronic camera 101 is the same as that described with reference to FIG. In the present embodiment, the processing of the fixed-length compression processing unit 205 described above is different from the first embodiment and the second embodiment.

  FIG. 10 is a graph for explaining the principle of the present embodiment, in which the relationship between the pseudo-quantization step (QP2) and the compression ratio (CR) is drawn in a macro manner as compared with FIG. 4 (c). That is, the lower limit value of the pseudo quantization step (QP2) on the horizontal axis in FIG. 4C is 1, and the upper limit value of the compression ratio (CR) on the vertical axis is 1, whereas in FIG. 1 or less of the lower limit value of the quasi-quantization step (QP2) of the axis and 1 or more of the upper limit value of the compression ratio (CR) of the ordinate are drawn macroscopically.

  The principle of the present embodiment is that, as shown in FIG. 10, when various types of images are compressed, the relationship between the pseudo quantization step (QP2) and the compression rate (CR) is the pseudo quantization step (QP2). Is assumed to be converged to one point in a region where the compression ratio (CR) is 1 or more. This principle is a concept obtained empirically based on experimental results by the inventors. For example, looking at the actually measured graph of FIG. 4 (c), it is easy to focus on one point in the upper left because it has a characteristic of spreading radially from the upper left to the lower right. Predictable.

  In the first embodiment and the second embodiment described above, it is necessary to select two quantization indexes for test compression, but in this embodiment, only one quantization index for test compression is selected. The advantage is that it can be selected and the test compression can be done only once.

  Hereinafter, a specific processing flow of the fixed-length compression processing unit 205 in the third embodiment will be described in order according to the flowchart of FIG. In the flowchart of FIG. 11, the processes with the same step numbers as those in the flowchart of FIG. 6 of the first embodiment are the same processes, and thus redundant description is omitted, and only the parts different from the flowchart of FIG.

In FIG. 11, the processing of step S101 and step S102 is the same as the flowchart of FIG.
(Step S203) The compression rate examining unit 262 first selects one test compression quantization index (QPIndex_1) corresponding to the target compression rate (CR_T) with reference to the table stored in the memory 109. As described above, the selection method selects the quantization index value having the highest probability.
(Step S204) The compression rate examining unit 262 gives the quantization index (QPIndex_1) for test compression to the quantization unit 252 of the compression processing unit 204, and performs test compression on the image data input to the compression processing unit 204. Do. Then, as in the first embodiment, the compression ratio (CR_1) corresponding to the quantization index (QPIndex_1) is obtained using (Equation 1).
(Step S205) The pseudo quantization width calculation unit 263 obtains a pseudo quantization step (QP2_1) corresponding to one quantization index (QPIndex_1) for test compression. On the other hand, as described with reference to FIG. 10, the initial pseudo-quantization step (QP2_0) and the initial compression rate (CR_0) of the portion concentrated on one point are obtained.
(Step S206) The pseudo quantization width calculation unit 263 uses the pseudo quantization step (QP2_1), compression rate (CR_1), initial pseudo quantization step (QP2_0), and initial compression rate (CR_0) obtained by the test compression. The target pseudo quantization step (QP2_T) at the target compression rate (CR_T) is calculated. Even if the calculation method is macroscopically as shown in FIG. 10, the relationship between the compression rate (CR) and the pseudo quantization step (QP2) is regarded as a straight line in the log-log graph as described in FIG. Therefore, the target pseudo quantization step (QP2_T) with respect to the target compression rate (CR_T) can be obtained by applying a linear interpolation method (Formula 6).

  (Equation 6) is an equation corresponding to (Equation 4) described in the first embodiment, and the pseudo-quantization step (QP2_2) and the compression rate (CR_2) of (Equation 4) are respectively the initial pseudo-quantization. Only the step (QP2_0) and the initial compression rate (CR_0) are replaced. Since the processing after step S106 is the same as that of the first embodiment, a duplicate description is omitted.

  In this way, the electronic camera 101 according to the third embodiment can obtain the quantization index from the test compression result only once so that the set target compression rate is obtained. As a result, the volume of the compressed image data can be made substantially constant, so that the number of images that can be captured with the target compression rate selected according to the remaining capacity of the memory card 112a can be grasped. Usability of the camera 101 is improved.

(Fourth embodiment)
Next, a fourth embodiment of the image compression apparatus and the electronic camera 101 according to the present invention will be described. The configuration of the electronic camera 101 is the same as that described with reference to FIG. Unlike the other embodiments, the present embodiment is characterized in that processing is performed with logarithmic values (log (QP2)) unlike the other embodiments. The reason for this is that, as shown in FIG. 2, the pseudo quantization step (QP2) takes a very large value as the quantization index (QPIndex) increases, and the pseudo quantization step (QP2) is obtained. Since only the logarithmic value is used, it is not necessary to calculate the QP2 value. In this embodiment, paying attention to this point, processing with logarithmic values can reduce the increase in logarithmic conversion and related processing amount and memory capacity.

  Hereinafter, the processing of the fixed-length compression processing unit 205 in the present embodiment will be described, but the basic processing flow is the same as the flowchart of FIG. 6 described in the first embodiment, and will be described with reference to FIG.

In the flowchart of FIG. 6, the processing from step S101 to step S103 is exactly the same, the processing of step S104 is slightly different from the first embodiment, and processing is performed as in step S104b below.
(Step S104b) The pseudo quantization width calculation unit 263 obtains logarithmic pseudo quantization steps (log (QP2_1) and log (QP2_2)) corresponding to the two quantization indexes for test compression (QPIndex_1 and QPIndex_2). . In the first embodiment, QP2 is obtained using a graph or table showing the relationship between the quantization index (QPIndex) and the pseudo quantization step (QP2) in FIG. 7A. The logarithmic value log (QP2) is used as it is. Thus, by using the logarithmic value log (QP2) of QP2 instead of QP2, for example, as shown in FIG. 2, the value of QP2 is a huge value in a section where i is 10 or more, but log (QP2 ) Is a size that is easy to handle.

Here, in the present embodiment, the program (1) described in the first embodiment is processed as a program (3) shown below.
[Program (3)]
for (i = 0; i <16; i ++) {
if (QPIndex_x> = QPIndex [i] && QPIndex_x <= QPIndex [i + 1]) {
logQP2_x = (logQP2 [i + 1] -logQP2 [i]) / (QPIndex [i + 1] -QPIndex [i]) *
(QPIndex_x-QPIndex [i]) + logQP2 [i];
break;
}
}
In this way, the pseudo quantization width calculation unit 263 obtains logarithmic pseudo quantization indexes (log (QP2_1) and log (QP2_2)) corresponding to the two quantization indexes for test compression (QPIndex_1 and QPIndex_2). Ask.
(Step S105b) The pseudo-quantization width calculation unit 263 uses the (Equation 4) in the same manner as in Step S105 of the first embodiment, and uses a logarithmic target pseudo-quantization step (log (QP2_T) In this embodiment, since the logarithmic pseudo-quantization index (log (QP2_1) and log (QP2_2)) is obtained in step S104b, log (QP2_1) and log (QP2_1) It is only necessary to substitute for QP2_2), and logarithmic calculation is not necessary.
(Step S106b) The target quantization index calculation unit 265 calculates a target quantization index (QPIndex_T) corresponding to the logarithmic target pseudo quantization step (log (QP2_T)) calculated in step S105b.

  Here, in the first embodiment, the target quantization index (QPIndex_T) is set using a graph or table showing the relationship between the quantization index (QPIndex) and the pseudo quantization step (QP2) in FIG. In this embodiment, the logarithm value log (QP2) of QP2 is used as it is. By using the logarithmic value log (QP2) of QP2 instead of QP2, it is possible to calculate the logarithm value log (QP2) that is easy to handle instead of QP2 that is a huge value as in step S104b.

Here, in this embodiment, the program (2) described in the first embodiment is processed as the following program (4).
[Program (4)]
for (i = 0; i <16; i ++) {
if (logQP2_T> = logQP2 [i] && logQP2_T <= logQP2 [i + 1]) {
QPIndex_T = (QPIndex [i + 1] -QPIndex [i]) / (logQP2 [i + 1] -logQP2 [i]) *
(logQP2_T-logQP2 [i]) + QPIndex [i];
break;
}
}
In this way, the target quantization index calculation unit 265 calculates the target quantization index (QPIndex_T) corresponding to the logarithmic target pseudo quantization step (log (QP2_T)) calculated in step S105.

  Subsequent processes in step S107 and step S108 are the same as those in the first embodiment, and thus redundant description is omitted.

  As described above, the electronic camera 101 according to the fourth embodiment can accurately obtain the quantization index from the test compression result so that the set target compression rate is obtained. As a result, the volume of the compressed image data can be made substantially constant, so that the number of images that can be captured with the target compression rate selected according to the remaining capacity of the memory card 112a can be grasped. Usability of the camera 101 is improved. In particular, in the present embodiment, the pseudo-quantization step is processed logarithmically, thereby reducing logarithmic conversion and related processing amount and memory capacity.

  In the above description, the description has been made based on the first embodiment. However, since the feature of the present embodiment is that the pseudo-quantization step is used as a logarithmic value, the second embodiment and the third embodiment. The same applies to the above.

  As described above, the embodiments of the image compression apparatus and the electronic camera according to the present invention have been described. By providing a pseudo quantization step in which the relationship between the compression rate and the quantization step is close to a straight line in the log-log graph. The target quantization index for the target compression rate can be obtained with high accuracy. As a result, the image compression apparatus and the electronic camera according to the present invention can make the image file capacity after compression almost constant regardless of what kind of subject is photographed, and the number of images that can be photographed with the remaining capacity of the recording medium. Prediction can be performed and the usability of the electronic camera is improved.

  In each embodiment, the target quantization index is obtained so as to be close to the target compression rate. However, a lower limit value (QPIndex_L) of the target quantization index is set and the predicted target quantization index ( When QPIndex_T) has the following relationship, the lower limit value (QPIndex_L) of the target quantization index may be used regardless of the predicted target quantization index (QPIndex_T). This prevents the problem that the quantization index used for the main compression becomes an extremely small value even when the captured image is an image with a small number of high-frequency components and the compression rate is high.

  As described above, the image compression apparatus and the electronic camera according to the present invention have been described by way of examples in the respective embodiments, but can be implemented in various other forms without departing from the spirit or main features thereof. . Therefore, the above-described embodiment is merely an example in all respects and should not be interpreted in a limited manner. The present invention is defined by the claims, and the present invention is not limited to the text of the specification. Further, all modifications and changes belonging to the equivalent scope of the claims are within the scope of the present invention.

DESCRIPTION OF SYMBOLS 101 ... Electronic camera 102 ... Optical system 103 ... Mechanical shutter 104 ... Image sensor 105 ... A / D conversion part 106 ... Image buffer 107 ... Image processing part 108 ... Camera control unit 109 ... Memory 110 ... Display unit 111 ... Operation member 112 ... Memory card IF (interface)
112a ... Memory card 201 ... White balance processing unit 202 ... Color interpolation processing unit 203 ... Edge enhancement processing unit 204 ... Compression processing unit 205 ... Fixed length compression processing unit 251 ... Orthogonal transformation unit 252 ... quantization unit 253 ... entropy coding unit 261 ... target compression rate setting unit 262 ... compression rate investigation unit (capacity investigation unit)
263 ... pseudo quantization width calculation unit 264 ... target pseudo quantization width calculation unit 265 ... target quantization index calculation unit

Claims (13)

Image data input means for inputting image data to be compressed;
First, the image data has a quantized change amount that changes almost linearly in relation to the file size, and a file size corresponding to the compression target image data is obtained for two predetermined quantized change amounts. One computing means;
A file size specification means for specifying a target file size;
Second computing means for obtaining a target quantization change amount corresponding to the target file size for the compression target image data based on the relationship between the file size obtained by the first computation means and the quantization change amount;
Storage means for storing in advance the relationship between the quantization change amount and the quantization step;
From the relationship between the quantization change amount stored in the storage means and the quantization step, a quantization step corresponding to the target file size is obtained based on the target quantization change amount obtained by the second calculation means. Image compression means for compressing the image data to be compressed using the obtained quantization step;
An image compression apparatus comprising: The image compression apparatus according to claim 1.
2. The image compression apparatus according to claim 1, wherein the two predetermined quantization change amounts are set according to the target file size. The image compression apparatus according to claim 1.
The first computing means obtains a file size for one quantized change amount set according to the target file size, and another quantized change amount determined according to the obtained file size. An image compression apparatus for obtaining a file size for. Image data input means for inputting image data to be compressed;
Storage means for storing a correspondence relationship between the quantization index and the quantization change amount;
Compression means for compressing the image data with the specified quantization index;
A file size specification means for specifying a target file size;
Test compression means for obtaining a first file size and a second file size when the image data is compressed by using the first quantization index and the second quantization index, respectively, by the compression means. When,
Using the correspondence relationship between the quantization index and the quantization change amount, the first quantization change amount corresponding to the first quantization index and the second quantization index corresponding to the second quantization index. A quantization variation calculation means for obtaining a quantization variation of
The relationship between the quantized change amount of the image data and the file size is represented by the first quantized change amount, the first file size, the second quantized change amount, and the second size. A target quantization change amount calculating means for obtaining a target quantization change amount at the target file size according to the obtained relationship.
A target quantization index calculating means for obtaining a target quantization index from the target quantization change amount using a correspondence relationship between the quantization index and the quantization change amount;
An image data output device comprising: image data output means for outputting compressed image data obtained by compressing the image data by the compression means at the target quantization index. The image compression apparatus according to claim 4.
The image compression apparatus according to claim 1, wherein the first quantization index and the second quantization index are set according to the target file size. The image compression apparatus according to claim 4.
The test compression unit obtains a first file size by performing a first compression by the compression unit using a first quantization index corresponding to the target file size, and according to the first file size An image compression apparatus characterized in that the second file size is obtained by performing a second compression by the compression means using the second quantization index determined in the above. The image compression apparatus according to any one of claims 4 to 6,
The image is characterized in that the first quantized change amount and the first file size and the logarithmic value of the second quantized change amount and the second file size are substantially straight lines. Compression device. The image compression apparatus according to any one of claims 4 to 7,
The quantization change amount calculation means calculates a logarithmic value of each first quantization change amount and a second quantization change amount from the first quantization index and the second quantization index. Logarithm value and
The target quantization change amount calculating means includes a logarithmic value of the first quantization change amount, a logarithmic value of the second quantization change amount, the first file size, and the second file size. The logarithmic value of the target quantization change amount at the target file size is obtained from the relationship with
The target quantization index calculating unit obtains a target quantization index from a logarithmic value of the target quantization change amount. Image data input means for inputting image data to be compressed;
Storage means for storing a correspondence relationship between the quantization index and the quantization change amount;
Compression means for compressing the image data with the specified quantization index;
Target file size setting means for setting a target file size of the compression means;
A test compression means for obtaining a first file size when the image data is compressed with a first quantization index by the compression means;
Using a correspondence relationship between the quantization index and the quantization change amount, a quantization change amount calculating unit for obtaining a first quantization change amount corresponding to the first quantization index;
The relationship between the quantized change amount of the image data and the file size, the first quantized change amount and the first file size, and a preset initial quantized change amount, A target quantized change amount calculating means for determining a target quantized change amount at the target file size according to a relationship obtained from an initial file size corresponding to the initial quantized change amount;
A target quantization index calculating means for obtaining a target quantization index from the target quantization change amount using a correspondence relationship between the quantization index and the quantization change amount;
An image data output device comprising: image data output means for outputting compressed image data obtained by compressing the image data by the compression means at the target quantization index. The image compression apparatus according to claim 9.
The image compression apparatus according to claim 1, wherein the first quantization index is set according to the target file size. The image compression apparatus according to claim 9 or 10,
The first compression change amount and the first file size, and the logarithmic value of the initial quantization change amount and the logarithmic value of the initial file size are substantially straight lines. apparatus. The image compression apparatus according to any one of claims 9 to 11,
The quantization change amount calculating means obtains a logarithmic value of the first quantization change amount from the first quantization index,
The target quantization change amount calculation means includes a logarithmic value of the first quantization change amount, a logarithm value of a preset initial quantization change amount, the first file size, and the initial quantization value. Obtain the logarithmic value of the target quantized change amount at the target file size from the relationship with the initial file size corresponding to the logarithmic value of the change amount,
The target compression index calculating unit obtains a target quantization index from a logarithmic value of the target quantization change amount.   An electronic camera comprising the image compression apparatus according to claim 1.

Images