JP5457853B2 - Image encoding apparatus, image encoding method, program, and storage medium - Google Patents

Description

  The present invention relates to an image encoding technique.

  Conventionally, there have been various methods for encoding an image with a fixed length. For example, Patent Literature 1 and Patent Literature 2 describe a method of separating image data into resolution information representing high-frequency components such as characters and lines and gradation information representing other components, and encoding each information. Is disclosed. Patent Document 3 discloses a method of encoding an image with a fixed length by binarizing the image and reducing the amount of identification information indicating which value each pixel has according to a predetermined rule. ing.

  On the other hand, the optimum encoding method differs depending on the characteristics of the image to be encoded. In particular, when an image is divided into a plurality of blocks and encoding is performed on each block, an optimum result cannot be obtained even if the same encoding method is applied to each block in the image. For example, Patent Document 1 discloses a technique in which each pixel is divided into a pixel group having a large pixel value and a pixel group having a small pixel value, and an encoding process is selected according to the difference in the average pixel value of each pixel group. Is disclosed.

JP 2008-278042 A JP 2009-17502 A JP 2009-5017 A

  In order to perform encoding suitable for the characteristics of an image, it is required to develop various selection techniques for encoding processing. By providing not only a method for selecting an encoding process according to the difference between the average pixel values of each pixel group as in Patent Document 1, but also a method for selecting an encoding process according to other characteristics of the image, The selection technique of the encoding process can be made more sophisticated.

  An object of the present invention is to provide a technique for selecting an appropriate encoding method according to the characteristics of an image.

In order to achieve the object of the present invention, for example, an image encoding apparatus of the present invention comprises the following arrangement. That is,
An image encoding device that encodes an image and outputs fixed-length encoded data including the encoded image,
An input means for acquiring a processing target image;
Generation means for generating, in the processing target image, a pixel whose pixel value is not less than a first threshold value and not more than a second threshold value, and generating position information indicating a position of the target pixel in the processing target image When,
Position encoding means for reversibly encoding the position information to generate encoded position data;
When the information amount of the encoded position data is equal to or less than a third threshold, first encoded gradation data having a first code amount is generated from the processing target image, and the first encoding level is generated. First gradation encoding means for generating pixel value information used for correcting the pixel value of the target pixel in an image obtained by decoding the key data;
When the information amount of the encoded position data is equal to or smaller than the third threshold, fixed-length encoded data including the first encoded gradation data, the pixel value information, and the encoded position data First output means for outputting;
Second gradation encoding means for generating second encoded gradation data having a second code amount from the processing target image when the information amount of the encoded position data exceeds the third threshold value When,
A second output means for outputting fixed-length encoded data including the second encoded gradation data when the information amount of the encoded position data exceeds the third threshold;
The first code amount and the second code amount are determined so that the fixed-length encoded data has a predetermined fixed length, and the second code amount is larger than the first code amount. To do.

  An appropriate encoding method can be selected according to the characteristics of the image.

1 is a block diagram illustrating an example of an image encoding device according to Embodiment 1. FIG. The flowchart which shows the detail of step S603. The figure which shows an example of the pattern of a semi-transparent part. The figure which shows an example of the histogram of a pixel value. FIG. 3 is a diagram illustrating a fixed-length code format according to the first embodiment. 3 is a flowchart according to the process of the first embodiment. FIG. 10 is a block diagram illustrating an example of a configuration of a computer according to a third embodiment.

Embodiments of the present invention will be described below with reference to the drawings. However, the scope of the present invention is not limited to the following examples.
[Example 1]
FIG. 1 is a block diagram illustrating an example of an image encoding apparatus according to the present embodiment. The blocking unit 101 acquires an input image and divides it into blocks having a predetermined size. Hereinafter, an image divided into blocks is referred to as a block image (processing target image). In the present embodiment, for the sake of explanation, each pixel of the block image has one pixel value. For example, when an input image is input as an RGB image, an R-plane block image, a G-plane block image, and a B-plane block image may be generated. As will be apparent to those skilled in the art, the block image may be an RGB image, but details are omitted for ease of explanation. In this embodiment, the block size is 16 × 16, the pixel value data width is 8 bits, the maximum pixel value included in each block is Cmax, and the minimum pixel value is Cmin.

  The extraction unit 102 separates each pixel of one block image into an extracted pixel and a non-extracted pixel (generation unit). The separation method will be described later. Then, position data indicating the position of the extracted pixel in the block image (in the processing target image) is generated. For example, the position data may be generated as a binary image having the same size as the block image, that is, 16 × 16 pixels. The pixel value of the binary image may be “1” if the corresponding pixel in the block image is an extracted pixel, and may be “0” if the corresponding pixel in the block image is a non-extracted pixel. In the present embodiment, since the position data is losslessly encoded, the position of the extracted pixel is well preserved after decoding.

  The position encoding unit 103 encodes the position data generated by the extraction unit 102 (position encoding unit). The position encoder 103 may be an encoder that performs lossless compression of position data. The encoding performed by the position encoding unit 103 may be variable length encoding. In the present embodiment, since the position encoding unit 103 compresses binary block data, a known encoding method such as run length encoding or Huffman encoding can be used. Further, the position encoding unit 103 performs encoding using the dichroic unit 104 by determining whether or not the generated encoded position data exceeds a predetermined threshold value, or replaces the replacement unit 105 and the gradation. It is determined whether to perform encoding using the encoding unit 106. This process will be described later.

  The dichroic unit 104 performs dichroization on the block image, encodes the dichroic block image, and generates encoded dichroic data (second encoded gradation data) (second encoding data) Gradation encoding means). The replacement unit 105 changes the pixel value of the extracted pixel according to the processing result of the extraction unit 102, and generates a gradation image in which the high-frequency component of the block image is suppressed. The processing of each of these units will be described later.

  The gradation encoding unit 106 encodes the gradation image to generate encoded gradation data (first encoded gradation data) (first gradation encoding means). The gradation encoding unit 106 may perform encoding using a known orthogonal transform such as JPEG in order to compress a multi-value block image. The output unit 107 generates fixed-length encoded data 108 using the encoded position data generated by the position encoder 103 and the encoded gradation data generated by the gradation encoder 106 (first). 2 output means). The output unit 109 generates fixed-length encoded data 110 using the dichroic data generated by the dichroic unit 104 or the encoded gradation data generated by the gradation encoding unit 106 ( First output means). A set of the fixed-length encoded data 108 and the fixed-length encoded data 110 output from the output unit 107 and the output unit 109 is aligned by an output unit (not shown) and output as the encoding result of the input image 100 (third). Output means). The processing of the output unit 107 and the output unit 109 will be described later.

  Here, the encoding method used in the present embodiment will be described. In this embodiment, four types of encoding methods are used: a difference mode, a replacement mode, a gradation mode, and a dichroic mode. However, some encoding methods may not be used. For example, only the dichroic mode and the gradation mode may be used. Below, each encoding method is demonstrated in detail.

(1) Differential mode The data format is shown in FIG. The encoded data in the differential mode includes a header, extracted color data (pixel value information), encoded position data, and encoded gradation data. In this mode, the gradation image output from the replacement unit 105 is an image obtained by subtracting a certain value from the pixel value of the extracted pixel included in the block image. The extracted color data shows this constant value. This constant value may be a difference between the representative pixel value of the extracted pixel and the representative pixel value of the non-extracted pixel. For example, this constant value may be a difference between the average value of the pixel values of the extracted pixels and the average value of the pixel values of the non-extracted pixels. This extracted color data is used to correct the pixel value of the extracted pixel in the image obtained by decoding the encoded gradation data. The extracted color data may indicate a negative value. The header can indicate that this data is encoded in differential mode.

(2) Replacement mode The data format is shown in FIG. The encoded data in the replacement mode includes a header, extracted color data, encoded position data, and encoded gradation data. In this mode, the gradation image output from the replacement unit 105 is an image obtained by setting the pixel value of the extracted pixel included in the block image to a constant value. This fixed value may be a representative pixel value of non-extracted pixels, for example, an average value. This extracted color data is used to correct the pixel value of the extracted pixel in the image obtained by decoding the encoded gradation data. In the replacement mode, the extracted color data indicates the pixel value of the extracted pixel after decoding. The extracted color data may be a representative pixel value of the extracted pixels, for example, an average value. The header can indicate that this data is encoded in replacement mode.

(3) Gradation mode The data format is shown in FIG. The encoded data in the gradation mode includes a header and compressed gradation data. In this mode, the replacement unit 105 outputs the block image as it is as a gradation image. The header can indicate that this data is encoded in grayscale mode.

(4) Two-color mode The data format is shown in FIG. The encoded data in the dichroic mode includes a header and encoded dichroic data. The header can indicate that this data is encoded in dichroic mode. Known methods can be used for dichroism and encoding. For example, encoding may be performed by a method such as that disclosed in Patent Document 3. As one example, the encoded dichroic data may include position information indicating which of the two pixel groups each pixel is included in and pixel values after decoding of the pixels included in each pixel group. Good.

  Next, processing in the present embodiment will be described with reference to the flowchart of FIG. In step S <b> 601, the blocking unit 101 acquires image data 100. In step S602, the blocking unit 101 divides the image data 100 into block images. In step S603, the extraction unit 102 separates each pixel of the block image into an extracted pixel and a non-extracted pixel. For example, the extraction unit 102 determines a lower limit TH0 (first threshold value) and an upper limit TH1 (second threshold value) (where TH0 ≦ TH1) of the pixel value range of the pixel to be extracted. Then, the extraction unit 102 may determine a pixel whose pixel value is TH0 or more (first threshold or more) and TH1 or less (second threshold or less) as an extraction pixel (target pixel).

  An example of a method for determining the range of pixel values of the extracted pixels by the extraction unit 102 will be described with reference to the flowchart of FIG. For example, when TH0 is Cmin, a pixel having a pixel value equal to or less than TH1 is an extraction pixel. When TH1 is Cmax, a pixel having a pixel value equal to or higher than TH0 is an extraction pixel. In step S201, the extraction unit 102 checks whether the pixel values of the pixels of the block image are the same. When each pixel value is the same, that is, when each pixel has the same color, the extraction unit 102 extracts all the pixels. Therefore, the extraction unit 102 sets (TH0, TH1) = (Cmax, Cmax).

  If the pixel values of the block image are not the same, the process proceeds to step S202. In step S202, the extraction unit 102 determines whether there are two types of pixel values of each pixel of the block image. When there are two types of pixel values, that is, when the block image is composed of two colors, the extraction unit 102 uses a pixel having a larger pixel value as an extraction pixel. Therefore, the extraction unit 102 sets the threshold value (TH0, TH1) = (Cmax, Cmax).

  If the pixel value is not two types, the process proceeds to step S203. In step S203, the extraction unit 102 determines whether a translucent part is included in the block image. The translucent portion is a regular pattern as shown in FIG. For example, such a pattern appears when a translucent color is used in a drawing application. In FIG. 3, black pixels indicate a semi-transparent color. Usually, the translucent color is composed of one color. On the other hand, white pixels represent the background portion. When a pattern as shown in FIG. 3 is detected, a semi-transparent color, that is, a black pixel in FIG.

  The determination in step S203 can be performed as follows. For example, when the pixel values of 64 pixels whose coordinates (x, y) are (2M, 2N) (where M and N are integers of 0 or more and less than 8) are all the same, the extraction unit 102 sets (C0, C0) as a threshold value. (TH0, TH1). At this time, C0 is a pixel value of coordinates (2M, 2N). On the other hand, if all the pixel values of the coordinates (2M, 2N) are not the same, and if all the pixel values of the coordinates (2M + 1, 2N) are the same, the extraction unit 102 sets (C0, C0) as the threshold value (TH0). , TH1). At this time, C0 is a pixel value of coordinates (2M + 1, 2N). Similarly, if not all the pixel values of the coordinates (2M + 1, 2N) are the same, the extraction unit 102 uses the pixel value as a threshold if the pixel values of the coordinates (2M, 2N + 1) are the same. When the pixel value of the pixel at the coordinate (2M, 2N + 1) is not common, if the pixel value of the pixel at the coordinate (2M + 1, 2N + 1) is the same, the pixel value is set as a threshold value. If the pixel values of the coordinates (2M + 1, 2N + 1) are not the same, the process proceeds to step S204.

  In step S204, the extraction unit 102 examines a pixel value histogram of the block image. When the extraction unit 102 determines that the appearance frequency of pixel values in the block image is uniform, all pixels are extracted pixels. An example of specific processing is shown below. The extraction unit 102 classifies each pixel into eight according to the pixel value. That is, the extraction unit 102 classifies pixels having pixel values of 0 or more and less than 32 into group 1, classifies pixels with pixel values of 32 or more and less than 64 into group 2, and similarly classifies the pixels into groups 3-8. The pixel values belonging to the group 8 are 224 or more and less than 256. If the number of groups in which the pixels are present is equal to or greater than the threshold (TH2) among the eight groups, the extraction unit 102 determines that the appearance frequency of the pixel values is uniform and sets all the pixels as extraction pixels. In this case, the extraction unit 102 may set (Cmin, Cmax) to (TH0, TH1). The range of TH2 is a value of 0 or more and 8 or less, but TH2 may normally be about 7.

  If the extraction unit 102 determines that the appearance frequency of pixel values is not uniform in step S204, the process proceeds to step S205. In step S205, the extraction unit 102 classifies each pixel into a pixel having a large pixel value and a pixel having a small pixel value, and sets one of the pixels as a selected pixel. An example of specific processing is shown below. The extraction unit 102 examines the histogram of each pixel value of the block image as in step S204. At this time, as shown in FIG. 4, the extraction unit 102 divides the range of pixel values of pixels constituting the block image into a plurality of areas, for example, 32 areas, and examines the appearance frequency of the pixel values included in each area. That is, the pixel value of the region 0 is 0 or more and less than 8, the pixel value of the region 1 is 8 or more and less than 16, and the pixel values entering the region 2 to the region 31 are similarly determined. The pixel value in the region 31 is 248 or more and less than 256.

  Then, the extraction unit 102 identifies a region (a non-pixel region) in which no pixel having a pixel value in the region (in a region) exists, and groups consecutive non-pixel regions. Of the grouped regions, the extraction unit 102 determines a region that does not include both the region 0 and the region 31 and has the maximum section length as the longest section (maximum length section). In the example of FIG. 4A, the areas 20 to 23 are the longest sections. The areas 27 to 31 are longer, but since the area 31 is included, it is not the longest section.

  Here, the range of pixel values in the longest section is (Lmin, Lmax). The minimum pixel value within the longest section (within the maximum length section) is Lmin, and the maximum pixel value is Lmax. In the example of FIG. 4A, (160, 191) is (Lmin, Lmax). Thus, the extraction unit 102 classifies the pixels into pixels having a pixel value larger than Lmax and pixels having a pixel value smaller than Lmin. On the other hand, there may be a case where the longest section does not exist. In this case, the extraction unit 102 may set all the pixels as extraction pixels. As a specific example, the extraction unit 102 may set (Cmin, Cmax) to (TH0, TH1).

  In step S205, the extraction unit 102 divides a section between the maximum value and the minimum value of the pixel values of the pixels in the block image (within the pixel value range of the pixels constituting the block image) into a plurality of regions. May be. In this case, the region having the maximum pixel value or the region having the minimum pixel value, for example, the region 0 or the region 31, always includes pixels. Therefore, since these areas are not included in the longest section, the longest continuous area may be determined as the longest section.

  In step S <b> 206, the extraction unit 102 determines which of the pixel having the large pixel value and the pixel having the small pixel value is the extraction pixel. Specific examples are shown below. First, the extraction unit 102 compares (Cmax−Lmax) (first difference value) with (Lmin−Cmin) (second difference value). When (Cmax−Lmax) is smaller, the extraction unit 102 sets a pixel having a large pixel value, that is, a pixel having a pixel value larger than Lmax as an extraction pixel. In this case, the extraction unit 102 may set (Lmax, Cmax) to (TH0, TH1). On the other hand, when (Lmin−Cmin) is smaller, the extraction unit 102 may select a pixel having a small pixel value, that is, a pixel having a pixel value smaller than Lmin as the extraction pixel. In this case, the extraction unit 102 may set (Cmin, Lmin) to (TH0, TH1).

  In steps S204 and S205, the extraction unit 102 determines whether or not there is a pixel in the region or group. However, it may be determined whether or not the number of pixels present in the region or group is greater than a predetermined value. Through the above steps S201 to S206, the extraction unit 102 can determine the pixel value range (TH0, TH1) of the extracted pixels. The extraction unit 102 may generate position data as described above using pixels whose pixel values fall within this range as extraction pixels.

  In this embodiment, the extraction unit 102 determines the extraction pixel by performing the processes of steps S201 to S206, but the extraction pixel determination method is not limited to this. In particular, it is not necessary to perform all the processes in steps S201 to S206. For example, only the process of step S204 may be performed. That is, when the extraction unit 102 finds a common pixel value C0, (C0, C0) may be set as the threshold value (TH0, TH1), and in all other cases, all pixels may be extracted.

  In step S604, the position encoding unit 103 encodes the position data generated in step S603 as described above. In step S605, the position encoding unit 103 determines whether the information amount of the encoded position data generated in step S604 exceeds a predetermined threshold (TH4) (third threshold). If the information amount of the encoded position data exceeds a predetermined threshold, the process proceeds to step S606, and the dichroic unit 104 encodes the block image into encoded dichroic data. That is, encoding in the dichroic mode is performed. If the information amount of the encoded position data is equal to or less than the predetermined threshold (TH4) (less than the third threshold), the process proceeds to step S607.

  For example, when the arrangement of extracted pixels is dispersed, the amount of information may not be reduced even if position data is encoded. In such a case, the output encoded data including the losslessly encoded position data may exceed a predetermined fixed length. In this embodiment, in such a case, encoding is performed in the dichroic mode, and position data is not included in the output encoded data. For example, the threshold value TH4 can be defined as fixed-length code amount− (header data amount + extracted color data amount + gradation data code amount minimum value). As an example, the fixed-length code amount is 256 bits, the header data amount is 3 bits, and the extracted color data amount is 9 bits. Further, the minimum value of the gradation data code amount is 20 bits so that the DC component is included in the gradation data. Then, TH4 becomes 224. For example, the minimum value of the gradation data code amount may be larger so that the gradation data includes the AC component, for example, so that the gradation data includes the AC component.

  In step S606, the dichroic unit 104 generates encoded dichroic data (second encoded gradation data) from the block image in the dichroic mode. The dichroic unit 104 binarizes the block image into the first pixel value and the second pixel value. For example, the dichroic unit 104 bisects the pixels in the block into a pixel group having a large pixel value and a pixel group having a small pixel value. Next, the dichroic unit 104 obtains a representative pixel value (for example, an average value of pixel values) for each pixel group. Then, the dichroic unit 104 uses, as encoded dichroic data, representative pixel values for each pixel group and position information (binary information) indicating which pixel group each pixel is included in. Output. At this time, the position information may be compressed.

  At this time, as in Patent Document 3, for example, the amount of position information can be reduced by changing pixel groups belonging to several pixels. In this case, the dichroic unit 104 may set the group of extracted pixels as the pixel group having the large pixel value or the pixel group having the small pixel value. In this case, the dichroic unit 104 may reduce the information amount of the position data generated by the extraction unit 102, for example, by reducing the number of extracted pixels according to a predetermined rule. Thus, the encoded dichroic data to be output can be set to a fixed length (second code amount).

  In step S607, the replacement unit 105 determines whether or not the difference between TH1 and TH0 obtained in step S603 is equal to or less than a predetermined value. If the difference between TH1 and TH0 is less than or equal to a predetermined threshold (TH3) (below the fourth threshold), the process proceeds to step S608 and encoding in the replacement mode is performed. TH1 indicates the maximum pixel value of the extracted pixel, and TH0 indicates the minimum pixel value of the extracted pixel. Therefore, when the pixel values of the extracted pixels are similar, the difference between TH1 and TH0 becomes small. In this embodiment, in this case, one pixel value close to the pixel value of the extracted pixel is determined, and a pixel value representative of this extracted pixel is used as the extracted color data. At the time of decoding, the pixel value indicated by the extracted color data may be the pixel value of the extracted pixel. In the present embodiment, encoding may be performed as the replacement mode when the difference between the maximum pixel value and the minimum pixel value of the extracted pixel is equal to or less than a predetermined threshold value.

  In step S608, the replacement unit 105 determines extracted color data indicating a pixel value (first extracted pixel value) similar to the pixel value of the extracted pixel. The value indicated by the extracted color data may be anything as long as it is a value (representative value) close to the pixel value of the extracted pixel. For example, the average value, the median value, or the mode value may be TH1. It may be an average with TH0. Furthermore, the value indicated by the extracted color data may be the maximum value or the minimum value of the pixel values of the extracted pixels.

  The threshold value TH3 may be constant, but the replacement unit 105 may determine the threshold value TH3. For example, when the pixel values of characters and lines vary, such as when the input image is read by a scanner, the replacement unit 105 may set the threshold value TH3 to a large value. On the other hand, when the density of characters and lines is uniform, for example, when the input image is created by a PC, the replacement unit 105 may set the threshold value TH3 to a value that is small or close to zero. In step S609, the replacement unit 105 generates a gradation image by replacing the pixel value of the extracted pixel with an appropriate value. For example, the replacement unit 105 may obtain a representative value (for example, an average value) of the pixel values of the non-extracted pixels and replace the pixel value of the extracted pixels with this representative value. The process in step S609 suppresses the high frequency component of the block image. But the process of step S609 is not essential for this invention. Even without the processing in step S609, the position of the extracted pixel can be stored with high accuracy.

  If the replacement unit 105 determines in step S607 that the difference between TH1 and TH0 obtained in step S603 is not less than a predetermined value, the process proceeds to step S610. In step S610, the replacement unit 105 determines whether all the pixels of the block image are extracted pixels. For example, when Cmin = TH0 and Cmax = TH1, all the pixels of the block image are extracted pixels. If all the pixels of the block image are extracted pixels, the process proceeds to step S611. In step S611, the replacement unit 105 transmits the block image as it is to the gradation encoding unit 106 as a gradation image. The gradation encoding unit 106 encodes the gradation image as described above, and sends the encoded gradation data to the output unit 107.

  When the extraction unit 102 determines in step S204 or step S206 that the appearance frequency of pixel values in the block image is uniform, all pixels in the block image are extracted pixels. For example, an input image with high gradation such as CZP (Circular Zone Plate) may have a locally complex shape. When such an input image is processed, the dichroic mode may be selected for a complicated block image. If the block image adjacent to the block image for which the dichroic mode is selected is encoded by classifying it into an extracted pixel and a non-extracted pixel, the change in image quality at the block boundary after decoding will be noticed. May be easier to stick. Therefore, when the appearance frequency of pixel values is uniform as in a high gradation image, the block image is compressed in the gradation mode.

  When the non-extracted pixel is included in the block image, the process proceeds from step S610 to step S612. In this case, encoding in the differential mode is performed. In step S612, the replacement unit 105 determines extracted color data indicating an appropriate pixel value. As described above, the extracted color data is the difference between the representative value (for example, the average value) of the pixel values of the extracted pixels and the representative value (for example, the average value) of the pixel values of the non-extracted pixels (second extracted pixel value). There may be. In step S613, the replacement unit 105 generates a gradation image by subtracting the pixel value indicated by the extracted color data from the pixel value of each extracted pixel. By this processing, the high frequency component of the block image is suppressed.

  In step S614, the gradation encoding unit 106 encodes the gradation image generated in step S609 (replacement mode) or step S613 (difference mode) as described above. In step S615, the output unit 107 outputs the fixed-length encoded data 108. In step S616, the output unit 109 outputs the fixed-length encoded data 110.

  An example of encoded data output by the output unit 107 and the output unit 109 is shown in FIG. When encoding is performed in the differential mode, the output unit 107 outputs the header, the extracted color data determined in step S612, the encoding position data generated in step S604, and the encoding generated in step S614. Outputs gradation data together. When encoding is performed in the replacement mode, the output unit 107 outputs the header, the extracted color data determined in step S608, the encoding position data generated in step S604, and the encoding level generated in step S614. Outputs key data together. The output unit 107 may stop the encoded gradation data so that the output data has a fixed length, or add bits to the encoded gradation data to obtain a predetermined code amount (first code amount). it can. However, the gradation encoding unit 106 may cut the data in advance or add bits so that the encoded gradation data has a predetermined code amount and the output data has a fixed length.

  When encoding is performed in the dichroic mode, the output unit 109 outputs the header and the encoded dichroic data generated in step S606 together. In the dichroic mode of this embodiment, encoded dichroic data is generated so that the output data has a fixed length. However, the output unit 107 may set the output data to a fixed length. When encoding is performed in the gradation mode, the output unit 109 outputs the header and the encoded gradation data generated in step S611 together. The output unit 109 can abort the encoded gradation data or add bits to the encoded gradation data so that the output data has a fixed length. However, the gradation coding unit 106 may cut off the data in advance or add bits so that the output data has a fixed length.

  According to this embodiment, the dichroic mode will be selected for high resolution block images. Also, the replacement mode will be selected for block images that contain distinct characters or lines. In addition, the gradation mode will be selected for block images that do not contain characters or lines. The difference mode will be applied to other block images. In this way, a compression mode suitable for each block image can be selected and encoded.

  In the present embodiment, the case where one of four encoding modes is selected has been described. However, it is also possible to select one from less than 3 or more than 5 encoding modes. For example, the encoding mode may be selected from the dichroic mode and the gradation mode. Also, the encoding mode may be selected from the dichroic mode, the difference mode, and the replacement mode. Furthermore, the encoding mode may be selected from the gradation mode and the difference mode. In this case, in step S606, encoding may be performed in the gradation mode instead of the dichroic mode. Also in these cases, selection can be performed according to the information amount of the encoded position data as in the present embodiment.

  In the present embodiment, when the code amount of the encoded position data is equal to or less than the threshold value, fixed-length encoded data including the encoded position data is output (difference mode, replacement mode). However, when all the pixels are extracted pixels, it is not necessary to include the encoded position data in the fixed-length encoded data (gradation mode). Further, when the code amount of the encoded position data exceeds the threshold value, the encoded position data is not included in the output fixed-length encoded data (dichroic mode). According to the configuration of this embodiment, when the encoded position data is included in the fixed-length encoded data and exceeds a predetermined fixed length, the fixed-length encoding is performed by an encoding method that does not use the encoded position data. be able to. Further, when the encoded position data is included in the fixed-length encoded data, it is possible to prevent the image from being deteriorated due to the encoded gradation data becoming too small.

[Example 2]
In this embodiment, the gradation image is reduced before the gradation encoding unit 106 encodes the gradation image. The image coding apparatus according to the present embodiment has a configuration shown in FIG. Regarding the processing according to the present embodiment, differences from the first embodiment will be described. In the present embodiment, the process is performed in step S651 before step S614. In step S651, the reduction unit 151 reduces the size of the gradation image from 16 × 16 to 8 × 8. Any reduction method may be used.

  For example, a method of reducing 4 pixels of 2 × 2 to one pixel can be used. In the difference mode or the gradation mode, the reduction unit 151 may set the average of the four pixel values as the pixel value after reduction. In the case of the replacement mode, the reduction unit 151 may refer to the position data and set the average pixel value of the non-extracted pixels as the pixel value after reduction. If all four pixels are extracted pixels, the pixel value of each pixel is common for the processing in step S609, and this common pixel value may be the reduced pixel value.

  It is generally important that character and line resolution information can have a high resolution in order to reproduce contours and shapes. However, the gradation information representing the background portion does not require as high a resolution as characters and lines. On the other hand, by reducing the gradation image, the data amount of the gradation image can be easily reduced. In this way, it is possible to reduce the code amount without easily perceiving a decrease in image quality. According to the present embodiment, it is possible to perform an encoding process suitable for human visual characteristics.

[Example 3]
In this embodiment, the processing according to each of the above-described embodiments is performed by a computer. FIG. 7 shows the basic configuration of the computer. In order to execute the functions of the above-described embodiments in this computer, each function may be expressed by a program and read into this computer. In this way, all the functions of the above-described embodiments can be realized by this computer. In this case, each of the components including FIG. 7 may be functioned by a function or a subroutine executed by the CPU. The computer program is usually stored in a computer-readable storage medium such as a CD-ROM. This storage medium can be executed by being set in a reading device (CD-ROM drive or the like) included in the computer and copied or installed in the system. Therefore, it is obvious that such a computer-readable storage medium is also within the scope of the present invention.

  In FIG. 7, the CPU 701 controls the operation of the entire computer. In this embodiment, the program is stored in a secondary storage 703 such as a hard disk or a CD-ROM. The CPU 701 reads a program into a primary storage 702 that is a memory such as a RAM, and executes the read program. The input device 704 is a device for inputting information to the computer, and corresponds to, for example, a mouse or a keyboard. The output device 705 is a device for outputting information by a computer, and includes a monitor and a printer. The reading device 706 is a device for acquiring external data, and includes a memory card reader and a network card. A bus 708 connects the above-described units to enable data exchange.

[Other Embodiments]
The present invention is also realized by executing the following processing. That is, software (program) that implements the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media. Then, the computer (or CPU, MPU, etc.) of the system or apparatus reads and executes the program code. In this case, the program and the storage medium storing the program constitute the present invention.

Claims (8)

An image encoding device that encodes an image and outputs fixed-length encoded data including the encoded image,
An input means for acquiring a processing target image;
Generation means for generating, in the processing target image, a pixel whose pixel value is not less than a first threshold value and not more than a second threshold value, and generating position information indicating a position of the target pixel in the processing target image When,
Position encoding means for reversibly encoding the position information to generate encoded position data;
When the information amount of the encoded position data is equal to or less than a third threshold, first encoded gradation data having a first code amount is generated from the processing target image, and the first encoding level is generated. First gradation encoding means for generating pixel value information used for correcting the pixel value of the target pixel in an image obtained by decoding the key data;
When the information amount of the encoded position data is equal to or smaller than the third threshold, fixed-length encoded data including the first encoded gradation data, the pixel value information, and the encoded position data First output means for outputting;
Second gradation encoding means for generating second encoded gradation data having a second code amount from the processing target image when the information amount of the encoded position data exceeds the third threshold value When,
A second output means for outputting fixed-length encoded data including the second encoded gradation data when the information amount of the encoded position data exceeds the third threshold;
The first code amount and the second code amount are determined so that the fixed-length encoded data has a predetermined fixed length, and the second code amount is larger than the first code amount. An image encoding device. A third output means;
The input means acquires an input image, acquires a plurality of the processing target images by dividing the input image,
The third output means outputs a set of fixed length encoded data output from the first or second output means for each of the processing target images as a result of encoding the input image. The image encoding apparatus according to claim 1, wherein: The first gradation encoding means includes:
Means for obtaining a first extracted pixel value that is a pixel value representative of the pixel value of each of the target pixels from the pixel value of each of the target pixels;
When the difference between the maximum value and the minimum value of the pixel values of each of the target pixels is equal to or less than a fourth threshold value, the pixel value information indicating the first extracted pixel value is generated, and the processing target image is used to generate the pixel value information. Means for generating first encoded gradation data;
When the difference between the maximum value and the minimum value is larger than the fourth threshold value, the pixel value of each pixel that is not the target pixel is represented by a pixel value that represents the pixel value of each pixel that is not the target pixel. A second extracted pixel value is obtained, the pixel value information indicating a difference between the first extracted pixel value and the second extracted pixel value is generated, and each of the target pixels in the processing target image A means for generating the first encoded gradation data by encoding an image obtained by subtracting a difference indicated by the pixel value information from a pixel value of The image encoding device according to 1 or 2. The generation means includes, among pixels constituting the processing target image,
When it is determined that the pixel values of all the pixels whose coordinates are (2M, 2N) (M and N are integers of 0 or more) are the same,
When it is determined that the pixel values of all the pixels whose coordinates are (2M + 1, 2N) are the same,
When it is determined that the pixel values of all the pixels whose coordinates are (2M, 2N + 1) are the same,
Or, when it is determined that the pixel values of all the pixels whose coordinates are (2M + 1, 2N + 1) are the same,
4. The image encoding device according to claim 1, wherein a pixel having a pixel value determined to be the same is the target pixel. 5. The generating means includes
Means for identifying, as a non-pixel section, a section in which no pixel having a pixel value in the section exists when dividing a range of pixel values of pixels constituting the processing target image into a plurality of sections;
Means for grouping sections in which the non-pixel sections are continuously arranged within the range, and specifying a group having a maximum section length as a maximum length section among the groups;
A first difference value obtained by subtracting the maximum pixel value in the maximum length section from the maximum pixel value in the range, and a first difference value obtained by subtracting the minimum pixel value in the range from the minimum pixel value in the maximum length section. Means for comparing the difference between the two difference values;
When the first difference value ≧ the second difference value, a pixel having a pixel value between the minimum pixel value in the maximum length section and the minimum pixel value in the range is set as the target pixel. Means,
When the first difference value <the second difference value, a pixel having a pixel value between the maximum pixel value in the maximum length section and the maximum pixel value in the range is set as the target pixel. Means,
The image encoding device according to any one of claims 1 to 3, further comprising: A method performed by an image encoding device for encoding an image and outputting fixed-length encoded data including the encoded image,
An input process for acquiring a processing target image;
In the processing target image, a generation step of generating, as a target pixel, a pixel whose pixel value is equal to or greater than a first threshold value and equal to or smaller than a second threshold value, and indicating position information of the target pixel in the processing target image When,
A position encoding step of reversibly encoding the position information to generate encoded position data;
When the information amount of the encoded position data is equal to or less than a third threshold, first encoded gradation data having a first code amount is generated from the processing target image, and the first encoding level is generated. A first gradation encoding step for generating pixel value information used for correcting a pixel value of the target pixel in an image obtained by decoding the key data;
When the information amount of the encoded position data is equal to or smaller than the third threshold, fixed-length encoded data including the first encoded gradation data, the pixel value information, and the encoded position data A first output step for outputting;
A second gradation encoding step of generating second encoded gradation data having a second code amount from the processing target image when the information amount of the encoded position data exceeds the third threshold value; When,
A second output step of outputting fixed-length encoded data including the second encoded gradation data when the information amount of the encoded position data exceeds the third threshold;
The first code amount and the second code amount are determined so that the fixed-length encoded data has a predetermined fixed length, and the second code amount is larger than the first code amount. An image encoding method.   The computer program for functioning a computer as each means which the image coding apparatus of any one of Claims 1 thru | or 5 has.   A computer-readable storage medium storing the computer program according to claim 7.

Images