JP3652754B2 - Subtractive color image processing device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a color-reduction image processing apparatus that performs color-reduction processing on image data, and more particularly to a color-reduction image processing apparatus that enables color reduction processing to be executed at high speed.
[0002]
The color-reduction image processing apparatus is prepared for realizing display of image data on a color display with a limited display color, and the pixel value of image data expressed using a large number of colors, The color reduction process is executed by replacing the closest representative color among the prepared representative colors.
[0003]
In order to improve the practicality of this color-reduced image processing apparatus, it is necessary to construct a configuration that enables the color-reduction process to be executed at high speed.
[0004]
[Prior art]
As a method for selecting a color closest to each pixel value of the original image from representative colors, there is a nearest neighbor search method. In this nearest neighbor search method, the color of each pixel of the original image is compared with all representative colors, and the representative color with the smallest difference, that is, the representative color with the smallest Euclidean distance in the color space is used as the replacement color. How to choose.
[0005]
Here, if an RGB color space as shown in FIG. 21 is assumed, the point x (R_x,G_x,B_x) And point y (R_y,G_y,B_y) Euclidean distance D (x, y) with
D (x, y) = [(R_x-R_y)²+ (G_x-G_y)²+ (B_x-B_y)²]^1/2
Or
D (x, y) = (R_x-R_y)²+ (G_x-G_y)²+ (B_x-B_y)²
Or
D (x, y) = | R_x-R_y| + | G_x-G_y| + | B_x-B_y｜
Are used.
[0006]
FIG. 22 illustrates the processing procedure of the nearest neighbor search method.
That is, in the nearest neighbor search method, as shown in this processing procedure, when the nth representative color p (n) is selected, the Euclidean distance D (s) between the representative color p (n) and the pixel value s of the original image. , p (n)), and the newly calculated Euclidean distance is the variable d_minWhen it is determined whether or not the distance is smaller than the smallest Euclidean distance calculated so far, the newly calculated Euclidean distance is set to the variable d._minAnd the n value at that time is stored in the variable n_minWhen the value of the variable n is incremented by one and then determined not to be small, the variable d_min/ N_minBy immediately repeating incrementing the value of the variable n by 1 without updating the color, the color closest to the pixel value of the original image is selected from the representative colors.
[0007]
In a subtractive color image processing apparatus that performs processing for selecting a color closest to each pixel value of an original image from representative colors using such a nearest neighbor search method, conventionally, for all pixel values of the original image, The search method was used to select and replace the closest color from the representative colors.
[0008]
That is, as shown in FIG. 23, when the pixel value to be processed is acquired from the original image, a replacement color is selected from the representative colors using the nearest neighbor search method, and the pixel value to be processed is After replacing with the replacement color, it is determined whether or not the processing of all the pixel values of the original image has been completed, and when it is determined that the processing has not been completed, it is repeated that the pixel value to be processed next is acquired. Thus, a configuration is adopted in which each pixel value of the original image is replaced with a representative color.
[0009]
[Problems to be solved by the invention]
However, according to such a conventional color-reduced image processing apparatus, there is a problem that it takes a long time to perform color-reduction processing of the original image. That is, in the nearest neighbor search method, since the distance between all the representative colors and each pixel value of the original image is calculated, the calculation amount becomes enormous, and it takes an enormous time for the color reduction processing of the original image. There was a problem.
[0010]
The present invention has been made in view of such circumstances, and an object of the present invention is to provide a new color-reduced image processing apparatus that can perform color-reduction processing at high speed.
[0011]
[Means for Solving the Problems]
FIG. 1 illustrates the principle configuration of the present invention.
In the figure, reference numeral 1 denotes a subtractive color image processing apparatus having the present invention, wherein pixel values of image data expressed using a large number of colors are replaced with the closest one of prepared representative colors. A color reduction process is executed.
[0012]
The reduced-color image processing apparatus 1 includes an original image file 10, a replacement image file 11, an input unit 12, a replacement unit 13, a cache memory unit 14, a determination unit 15, a conversion unit 16, and an activation unit 17. Is provided.
[0013]
  The original image file 10 stores an original image to be reduced. The replacement image file 11 stores the reduced color replacement image.
  The input unit 12 receives the pixel value from the original image file 10.One by oneRead and issue a replacement request. Here, the input means 12 forcibly sets a zero value to the lower bits of the read pixel value and then issues a replacement request, or forcibly sets an intermediate value of values that the lower bits can take. Then, a replacement request may be issued.
  Replacement means 13Calculates the distance between the pixel value for which there is a replacement request and a plurality of representative colors to be prepared, and based on the distance, replaces the pixel value with the closest one of the representative colors to be prepared, PlaceThe replacement color of the pixel value for which replacement is requested is obtained and replaced.
[0014]
  The cache memory means 14 determines the correspondence between the pixel value obtained by the replacement means 13 and the replacement color.Temporarilysave. Here, the cache memory unit 14 may store the correspondence using an index value indicating the replacement color, instead of storing the correspondence using the pixel value itself of the replacement color. The determination unit 15 determines whether or not a pixel value for which a replacement request has been made is registered in the cache memory unit 14. The conversion unit 16 registers the pixel value for which a replacement request is made in the cache memory unit 14.HaveTo the replacement color indicated by the pixel value. The activation unit 17 activates the replacement unit 13 by designating a pixel value for which a replacement request is made.
[0015]
  In the subtractive color image processing apparatus 1 of the present invention configured as described above, the input unit 12 divides the original image stored in the original image file 10 into a plurality of rectangular blocks, and uses the rectangular blocks as scanning units. Pixel valueOne by oneRead and issue a replacement request. Upon receiving this replacement request, the determination unit 15 refers to the cache memory unit 14 to determine whether or not a pixel value for which a replacement request has been made is registered in the cache memory unit 14 and is registered. Sometimes, the replacement color indicated by the pixel value is specified.
[0016]
In response to the processing of the determination unit 15, the conversion unit 16 converts the pixel value for which replacement is requested into the replacement color specified by the determination unit 15 and stores it in the replacement image file 11 when the determination unit 15 determines registration. To do. On the other hand, the activation unit 17 activates the replacement unit 13 by designating a pixel value for which a replacement request is made when the determination unit 15 determines that it is not registered.
[0017]
Then, the replacement unit 13, when activated by the activation unit 17, executes a replacement process by obtaining a replacement color of a pixel value having a replacement request using a nearest neighbor search method or the like, and stores it in the replacement image file 11. At the same time, the correspondence relationship between the obtained pixel value and the replacement color is registered in the cache memory means 14.
[0018]
Thus, in the present invention, when adopting a configuration for performing the color reduction processing using the nearest neighbor search method or the like, since the cache memory is used, the configuration for reducing the number of color reduction processing calculations is adopted. Multicolor image reduction can be performed at high speed.
[0019]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, the present invention will be described in detail according to embodiments.
FIG. 2 shows an embodiment of a subtractive color image processing apparatus 1 having the present invention.
[0020]
As shown in this figure, a reduced-color image processing apparatus 1 having the present invention includes an original image file 20 that stores an original image to be reduced, an image input program 21 that reads image values from the original image file 20, and an image A color reduction processing program 22 for obtaining a replacement color of the pixel value by performing a color reduction process on the pixel value read by the input program 21 and a representative color to be a replacement color prepared for the processing of the color reduction processing program 22 A color palette 23 to be managed, a cache mechanism 24 that is prepared for the processing of the color reduction processing program 22 and manages the correspondence between the pixel value obtained by the color reduction processing program 22 and the replacement color, and the color reduction processing program 22 A replacement image file 25 for storing a replacement image, and an image output program for reading image values from the replacement image file 25 and outputting them to the outside And a 6.
[0021]
FIG. 3 illustrates an example of management data of the color palette 23, and FIG. 4 illustrates an example of management data of the cache mechanism 24.
As shown in FIG. 3, the color palette 23 manages, for example, the correspondence between 256 prepared representative color index values and RGB values serving as pixel values of these representative colors.
[0022]
On the other hand, the cache mechanism 24 has, for example, M entries, and replaces the pixel value obtained by the color reduction processing program 22 while using the index value managed by the color palette 23 as shown in FIG. As shown in FIG. 4B, the pixel values and replacement colors obtained by the color reduction processing program 22 are managed while directly using the replacement color pixel values without using the index values, as shown in FIG. Manage the correspondence of. Here, of the two management configurations, the former management configuration using the index value is advantageous in that the memory can be greatly reduced.
[0023]
The cache mechanism 24 temporarily holds the correspondence relationship between the pixel value obtained by the color reduction processing program 22 and the replacement color. As will be described later, when the correspondence relationship is held, the correspondence referred to is referred to. A management configuration is adopted in which the new correspondence in the registration order is at the head while the relationship is at the head. This management configuration deletes the oldest registered correspondence managed in the last entry when a new correspondence is obtained by the color reduction processing program 22 while rearranging the referenced correspondence at the top. It is realized by registering the newly found correspondence relationship in the top entry while moving the correspondence relationship managed by the entry to the entry managing the old one by one. It takes time to move data.
[0024]
Therefore, the present invention employs a configuration that employs a link structure and rearranges the data by replacing the links without physically moving the data. That is, the cache mechanism 24, as shown in FIG. 5, corresponds to each entry, the management area prev (m) for managing the entry in which the previous correspondence is registered, and the next correspondence It has a link structure consisting of a management area next (m) that manages which entry the relationship is registered in, and can be referenced without moving data by re-linking those links. The correspondence relationship is managed in accordance with a format in which the correspondence relationship in the new registration order becomes the top side while rearranging the correspondence relationship thus formed.
[0025]
FIG. 6 shows an embodiment of a processing flow executed by the image-reduction image processing apparatus 1 of the present invention by starting the image input program 21 / color-reduction processing program 22.
As shown in the processing flow of FIG. 6, the color-reduced image processing apparatus 1 of the present invention first sets the reference position of the pixel to the first pixel of the original image in step 1. The current pixel value (the pixel value of the set reference pixel) is acquired from the file 20, and the acquired current pixel value is retrieved from the cache mechanism 24 in the subsequent step 3.
[0026]
In the subsequent step 4, it is determined whether or not the current pixel value is registered in the cache mechanism 24, and when it is determined that the current pixel value is not registered, the process proceeds to step 5 and the nearest neighbor search method is used. A replacement color is selected from the representative colors registered in the color palette 23, and in the subsequent step 6, the correspondence relationship between the obtained pixel value and the replacement color is registered in the cache mechanism 24.
[0027]
On the other hand, when determining that it is registered in step 4, go to step 7 to obtain a replacement color associated with the current pixel value from the registered correspondence, and in step 8 that follows, The relationship is moved to the top entry of the cache mechanism 24.
[0028]
When the replacement color is specified in Step 5 / Step 7, the process proceeds to Step 9, where the current pixel value is replaced with the replacement color, and then stored in the replacement image file 25. Then, in Step 10, it is stored in the original image file 20. When it is determined whether or not the processing has been completed for all the pixels, and it is determined that there are no remaining pixels, the processing is ended. After setting the reference position to the next pixel, the process returns to step 2.
[0029]
Thus, the color-reduction image processing apparatus 1 of the present invention employs a configuration that reduces the number of color-reduction processing calculations by using the cache mechanism 24.
The image input program 21 executes the processing of Step 1 / Step 2 / Step 10 / Step 11 in the processing flow of FIG.
[0030]
That is, as shown in the processing flow of FIG. 7, the image input program 21 first sets the input position at the top of the pixel in Step 1 and then in Step 2 receives an image input request from the color reduction processing program 22. When it is detected that there is an input request, the process proceeds to step 3 to determine whether or not the processing has been completed for all the pixels of the original image file 20, and when determining the end of all the pixels, step 4 is performed. Then, the end of input is notified to the color reduction processing program 22.
[0031]
On the other hand, when it is determined in step 3 that all the pixels have not been completed, the process proceeds to step 5 where a pixel value for 3 bytes is read from the set input position and is passed to the color reduction processing program 22. Is set 3 bytes later, and the process returns to step 2. Here, in step 5, the pixel value for 3 bytes is read from the input position and passed to the color reduction processing program 22 because the pixel value is represented by 3 bytes of RGB as shown in FIG. is there.
[0032]
As described above, the image input program 21 reads out the pixel values from the original image file 20 and passes them to the color reduction processing program 22. In order to improve the cache hit rate, the image input program 21 performs normal raster scan scanning. Instead of adopting a pixel value reading method, it is preferable to divide the original image into a plurality of rectangular blocks and read the pixel values using the rectangular blocks as scanning units.
[0033]
That is, in a normal image, pixels having the same color are concentrated in a relatively narrow range having a two-dimensional spread. Therefore, the image input program 21 performs normal raster scan scanning as shown in FIG. 9B, the original image as shown in FIG. 9B is divided into a plurality of rectangular blocks, and the pixel values are read out using the rectangular blocks as scanning units. It is preferable to adopt a configuration that realizes an improvement in the cache hit rate.
[0034]
FIG. 10 shows an example of a processing flow executed by the image input program 21 in order to realize the pixel value reading method shown in FIG. 9B.
Here, in this processing flow, as shown in FIG. 11, the original image has X pixels in the horizontal direction and Y pixels in the vertical direction with the upper left pixel as the origin (0, 0). In addition, as shown in FIG. 12, it is assumed that the image is partitioned by rectangular blocks having BX pixels in the horizontal direction and BY pixels in the vertical direction. Also, x is the pixel position in the horizontal direction of the readout target pixel, y is the pixel position in the vertical direction of the readout target pixel, bx is the horizontal block number of the rectangular block shown in FIG. 12, and by is the rectangle shown in FIG. It represents the block number in the vertical direction of the block, and these initial values are all zero. If the pixel value adopts a one-dimensional memory development form as shown in FIG. 8, the address A where the pixel value of the pixel coordinate (x, y) is developed._iIs "A_i= (X · y + x) · 3 ”.
[0035]
When the image input program 21 reads a pixel value from the original image according to the processing flow of FIG. 10, first, in step 1, the value of the variable x indicating the horizontal pixel position of the pixel to be read is incremented by one, In the subsequent step 2, it is determined whether or not the remainder (x% BX) obtained by dividing x by BX has become "0". When this remainder becomes “0”, it means that the pixel of the next rectangular block in the horizontal direction has been entered, so in this step 2, the pixel indicated by the variable x enters the rectangular block on the right. It is judged whether it becomes.
[0036]
When it is determined in step 2 that the rectangular block adjacent to the right has not been reached, the process proceeds to step 3 to determine whether or not the value of the variable x has reached X. That is, it is determined whether or not the final pixel position in the horizontal direction has been reached. When it is determined that the final pixel position in the horizontal direction has not been reached by this determination process, the value of the variable x updated in step 1 is treated as valid, and the pixels specified by the variable x and the variable y are read. As a target, the process of reading the pixel value of the pixel is started.
[0037]
On the other hand, when it is determined at step 2 that the right-hand side rectangular block is entered and when it is determined at step 3 that the final pixel position in the horizontal direction has been reached, the process proceeds to step 4 to set the vertical pixel position of the pixel to be read out. The value of the indicated variable y is incremented by 1, and in the subsequent step 5, it is determined whether or not the remainder (y% BY) obtained by dividing y by BY has become “0”. If this remainder becomes “0”, it means that the pixel of the next rectangular block in the vertical direction has been entered, so in this step 5, the pixel indicated by the variable y enters the lower rectangular block. It is judged whether it becomes.
[0038]
If it is determined in step 5 that the lower rectangular block has not been reached, the process proceeds to step 6 to determine whether the value of the variable y has reached Y. That is, it is determined whether or not the final pixel position in the vertical direction has been reached. When it is determined by the determination process in step 6 that the final pixel position in the vertical direction has not been reached, the process proceeds to step 7 where the multiplication value of the variable bx and the variable BX is substituted into the variable x. After setting the value of x so that it points to the position of the first pixel in the horizontal direction of the rectangular block, the process begins to read out the pixel value of that pixel with the pixel specified by the variable x and variable y as the read target.
[0039]
On the other hand, when it is determined at step 5 that the next lower rectangular block is entered and when it is determined at step 6 that the final pixel position in the vertical direction has been reached, the routine proceeds to step 8 where the value of the variable bx is incremented by one. In step 9, it is determined whether or not the conditional expression “bx × BX> X” is satisfied. If this conditional expression is satisfied, it means that the final rectangular block in the horizontal direction has been exceeded. In Step 9, it is determined whether or not such a state has been reached.
[0040]
If it is determined in step 9 that the conditional expression “bx × BX> X” is satisfied, the process proceeds to step 10 to increment the variable by indicating the vertical block number of the rectangular block to which the pixel to be read belongs by one. In the following step 11, the value of the variable bx indicating the horizontal block number of the rectangular block to which the pixel to be read belongs is set to “0”. Subsequently, in step 12, the product of the variable by and the variable BY is set to the variable y. Then, the process proceeds to step 7 and the process of reading out the pixel value of the pixel with the pixel designated by the variable x and the variable y as the reading target is started.
[0041]
On the other hand, when it is determined in this step 9 that the conditional expression “bx × BX> X” is not satisfied, the processing of step 10 and step 11 is omitted, and the process immediately proceeds to step 12, where the variables by and BY After the multiplication value is set to the variable y, the process proceeds to step 7 and the process of reading the pixel value of the pixel with the pixel specified by the variable x and the variable y as a reading target is started.
[0042]
In this way, the image input program 21 reads out pixel values from the original image file 20 in accordance with the scanning format shown in FIG. 9B by executing the processing flow of FIG. As a result, the cache hit rate can be improved.
[0043]
Next, processing executed by the color reduction processing program 22 will be described.
The subtractive color processing program 22 executes the processing from step 3 to step 9 in the processing flow of FIG.
[0044]
FIG. 13 illustrates an example of a processing flow executed by the color reduction processing program 22.
That is, when the color reduction processing program 22 is started, as shown in the processing flow of FIG. 13, first, in step 1, an image input request is issued to the image input program 21, and then in step 2. In response to the input request, it is determined whether or not the input end notification is notified from the image input program 21. When the input end notification is received, the process proceeds to step 2 to process the replacement image file 25. Notify the end.
[0045]
On the other hand, when it is determined in step 2 that what is received from the image input program 21 is not an input end notification, that is, when it is determined that what is received from the image input program 21 is a pixel value, the process proceeds to step 3. The pixel value is received, and in step 5 that follows, the cache mechanism 24 is searched using the pixel value.
[0046]
Subsequently, at step 6, it is determined whether or not there is a cache hit by the search processing at step 5, and when it is determined that there is no cache hit, the process proceeds to step 7 and the nearest neighbor search method is used to set the color palette. 23, the replacement color of the pixel value received in step 3 and its index value are selected from the representative colors registered in step 23, and in the subsequent step 8, the correspondence between the pixel value and the index value is determined as the top of the cache mechanism 24 Register in the entry.
[0047]
On the other hand, when it is determined in step 6 that a cache hit has occurred, the process proceeds to step 9 in which the replacement color indicated by the index value of the cache hit destination is specified. Move to the first entry.
[0048]
When the replacement color is specified in step 7 / step 9, the process proceeds to step 11 where an output request is issued to the replacement image file 25, and in step 12 the pixel value received in step 3 is replaced with the replacement color. Then, the data is stored in the replacement image file 25 (the replacement color itself may be stored or the index value may be stored), and then the process returns to Step 1.
[0049]
In this way, the color reduction processing program 22 uses the cache mechanism 24 to reduce the number of color reduction processing calculations.
Next, the configuration of the cache mechanism 24 will be described in detail.
[0050]
As shown in FIG. 5, the cache mechanism 24 manages each entry corresponding to the management area prev (m) for managing the entry corresponding to the previous correspondence and the next correspondence. It has a link structure consisting of a management area next (m) that manages which entry is registered, and by re-linking those links, it can be referenced without moving the data. The correspondence relationship is managed in accordance with a format in which the correspondence relationship in the new registration order becomes the top side while rearranging the correspondence relationship at the top.
[0051]
That is, when the correspondence relationship is not registered in the cache mechanism 24, as shown in FIG. 14A, Top indicating the first entry points to Last indicating the last entry. In this state, if there is a correspondence registration request, as shown in FIG. 14 (b), the correspondence with the registration request is registered in the entry No. 0 as the first entry, and An entry is set so that prev (m) indicates Top, next (m) indicates Last, and Top indicates a 0th entry, and Last indicates a 0th entry.
[0052]
Subsequently, when there is a correspondence registration request, the registration request correspondence is registered at the head of the free entry, and prev (m) of the storage destination entry indicates Top, and next (m) Is set to point to the entry that previously pointed to Top, and prev (m) of the entry that pointed to Top until then is set to point to the storage destination entry. Set to point.
[0053]
In this manner, as shown in FIGS. 15A and 15B, the correspondence with a registration request is registered in the cache mechanism 24 until the entry is full.
Subsequently, when there is a correspondence registration request, as shown in FIG. 16A, the correspondence with the registration request is registered in the entry indicated by Last, and prev (m) of the storage destination entry stores Top. And next (m) is set to point to the entry that previously pointed to Top, and prev (m) of the entry that previously pointed to Top is set to point to the storage destination entry. Next (m) of the entry pointed to by prev (m) of the storage destination entry is set to point to Last, and further, Top indicates the storage destination entry, and Last indicates prev () of the storage destination entry. Set to point to the entry pointed to by m).
[0054]
In this state, when the correspondence stored in the (M-1) th entry is referred to, as shown in FIG. 16 (b), the prev that the (M-1) th entry has (m) points to Top, next (m) is set to point to the entry that previously pointed to Top, and prev (m) of the entry that previously pointed to Top is number (M-1) Next (m) is set to point to the entry pointed to by next (m) of the (M-1) th entry, and Top points to the (M-1) th entry. Set as follows.
[0055]
In this way, the cache mechanism 24 has a link structure, and the links are rearranged, so that the registered correspondence is rearranged while rearranging the referenced correspondence relationship without moving the data. A configuration is adopted in which the correspondence relationship is managed in accordance with the format in which the correspondence relationship is the leading side.
[0056]
In accordance with this configuration, when the color reduction processing program 22 enters the cache search process in step 6 of the process flow shown in FIG. 13, as shown in the process flow of FIG. The index value to point to is searched. Here, in the figure, s represents a pixel value to be searched, c (m) represents a pixel value stored in the mth entry, and n (m) represents an index value stored in the mth entry. Yes.
[0057]
According to the processing of the color reduction processing program 22, the replacement image file 25 is replaced with the replacement color index value (the index value is represented by 1 byte because the color palette 23 manages 256 representative colors). Is stored, the image output program 26 executes a process of reading the image value from the replacement image file 25 and outputting it to the outside according to the processing flow shown in FIG.
[0058]
That is, as shown in the processing flow of FIG. 18, the image output program 26 first sets the output position at the top of the pixel in step 1 and issues an end notification from the replacement image file 25 in the subsequent step 2. If the end notification is issued, the process ends. If the end notification is not issued, subsequently, in step 3, whether or not a replacement image output request is issued from the output destination. If the output request issuance is detected, the process proceeds to step 4 to output the representative color indicated by the index value received from the replacement image file 25 to the output position. In the subsequent step 5, the output position is updated. By repeating the process from step 2 to step 2, the image value is read from the replacement image file 25 and output to the outside.
[0059]
As described above, when the image input program 21 reads out pixel values from the original image file 20 and passes them to the color reduction processing program 22, it reads out the pixel values according to the scanning format shown in FIG. The cache hit rate is improved, but in combination with or independently, the lower bit of the pixel value read from the original image file 20 is forcibly set to a zero value to thereby improve the cache hit rate. It is also possible to adopt a configuration that realizes an improvement in rate.
[0060]
Although the pixel value represented by 24 bits in total of 8 bits for each of RGB is about 16 million colors, such a large number of colors cannot be identified with the naked eye. Therefore, for example, even if each of the lower 4 bits of RGB is ignored and reduced to 4096 colors, there is no practical problem. Therefore, the image input program 21, for example, forcibly sets the lower 4 bits of the pixel value read from the original image file 20 to zero value, thereby reducing the color of the original image to 4096 colors, thereby It is possible to improve the hit rate.
[0061]
At this time, if the lower 4 bits are all zeros, the original value that can take a value between “1111” and “0000” is unified with “0000”, and the image becomes dark. Therefore, instead of forcibly setting the lower 4 bits to “0000”, as shown in FIG. 19, there is a method for forcibly setting “1000” which is an intermediate value from “1111” to “0000”. It is preferable to take.
[0062]
When such a method of reducing the lower bits is adopted, the subtractive color image processing apparatus 1 performs the original image processing between Step 2 and Step 3 of the processing flow of FIG. 6, as shown in Step 20 of the processing flow of FIG. By ignoring the low-order bits of the current pixel value acquired from the file 20, processing for reducing the number of bits is executed.
[0063]
Although the present invention has been described in detail according to the illustrated embodiments, the present invention is not limited thereto. For example, although pixel values represented in the RGB color system are used in the embodiments, the present invention is not limited to this. In the embodiment, the color reduction process is executed according to the nearest neighbor search method. However, the present invention is not limited to this.
[0064]
【The invention's effect】
As described above, the present invention employs a configuration for reducing the number of color reduction processing calculations by using a cache memory when adopting a configuration for executing the color reduction processing using the nearest neighbor search method or the like. As a result, the color reduction of the multicolor image can be executed at high speed.
[Brief description of the drawings]
FIG. 1 is a principle configuration diagram of the present invention.
FIG. 2 is an example of the present invention.
FIG. 3 is an example of color pallet management data.
FIG. 4 is an example of management data of a cache mechanism.
FIG. 5 is an example of management data of a cache mechanism.
FIG. 6 is an example of a processing flow executed by the present invention.
FIG. 7 is an example of a processing flow of an image input program.
FIG. 8 is an explanatory diagram of a storage form of pixel values.
FIG. 9 is an explanatory diagram of a pixel value reading method of an image input program.
FIG. 10 is an example of a processing flow of an image input program.
FIG. 11 is an explanatory diagram of a coordinate system of pixel values.
FIG. 12 is an explanatory diagram of a rectangular block.
FIG. 13 is an example of a processing flow of a color reduction processing program.
FIG. 14 is an operation explanatory diagram of a cache mechanism.
FIG. 15 is an operation explanatory diagram of a cache mechanism.
FIG. 16 is an operation explanatory diagram of a cache mechanism.
FIG. 17 is an example of a processing flow of a color reduction processing program.
FIG. 18 is an example of a processing flow of an image output program.
FIG. 19 is an explanatory diagram of processing executed by the image input program.
FIG. 20 is another embodiment of the processing flow executed by the present invention.
FIG. 21 is an explanatory diagram of a color space.
FIG. 22 is an explanatory diagram of a nearest neighbor search method.
FIG. 23 is an explanatory diagram of a prior art.
[Explanation of symbols]
1. Color reduction image processing device
10 Original image file
11 Replacement image file
12 Input means
13 Replacement means
14 Cache memory means
15 Judgment means
16 Conversion means
17 Starting means

Claims (7)

In the subtractive color image processing apparatus for performing color reduction processing of images data,
Input means for reading out and inputting pixel values one by one from image data to be subjected to color reduction processing;
Using the pixel value input by the input means as a replacement object, the distance between the pixel value and a plurality of representative colors prepared is calculated, and based on this, the pixel value is the closest of the representative colors Replacement means for replacing with,
Cache memory means for temporarily storing the correspondence between the pixel value obtained by the replacement means and the replacement color;
A determining means for image pixel value input of the input means to determine whether or not registered in the cache memory means,
If the determination means you determines registration, and converting means for replacing the image pixel value input of the input means to the replacement color indicated by the pixel value registered in said cache memory means,
If the determination means you determine unregistered, by specifying the image pixel value input of the input means further comprising a starting means for starting the replacement means,
A feature of a reduced color image processing apparatus. The color-reduced image processing apparatus according to claim 1.
It said cache memory means, according to the link structure, with the referenced correspondence to the head, to save the corresponding relationship between such a new correspondence between the registration order is top side,
A feature of a reduced color image processing apparatus. In the subtractive color image processing apparatus according to claim 1 or 2,
The cache memory means stores the correspondence relationship using the pixel value itself of the replacement color.
A feature of a reduced color image processing apparatus. In the subtractive color image processing apparatus according to claim 1 or 2,
The cache memory means stores the correspondence relationship using an index value indicating a replacement color.
A feature of a reduced color image processing apparatus. The color-reduced image processing apparatus according to claim 1 .
Said input means, and divides the images data to be color reduction process in a plurality of rectangular blocks, that it will read the pixel values that rectangular block as a scanning unit,
A feature of a reduced color image processing apparatus. In the color-reduced image processing apparatus according to claim 1 or 5 ,
Said input means, to the lower bits of the read pixel values, the set Tosu Rukoto forcibly zero value,
A feature of a reduced color image processing apparatus. In the color-reduced image processing apparatus according to claim 1 or 5 ,
Said input means, to the lower bits of the read pixel values, the set Tosu Rukoto intermediate values forcibly possible values of the lower significant bits,
A feature of a reduced color image processing apparatus.

Images