Error-diffusion image processing system and method thereof
View Patent ↗An improved error-diffusion image processing system and the method thereof adopt the block reading mechanism to replace the row reading mechanism employed in the conventional error-diffusion image processing. This can greatly reduce the memory space required for image processing and increase the operational efficiency. The problem of gaps existing in the block reading mechanism is solved by adding a required error sum (BEsum) in a summed pixel range according to specific position conditions of each block.
1. An error-diffusion image processing system to convert a multi-level image into a bi-level image, comprising:
an image partition unit, which divides the multi-level image into a plurality of blocks, each having a position condition (i,j) and a plurality of pixels;
an operation block processing unit, which reads in a block according to a processing order and determines an operation block range according to the position condition (i,j) to compute a summed error (BEsum) associated with the block;
a summed pixel processing unit, which determines a summed pixel range according to the position condition (i,j) of the block and adds the summed error to the summed pixel range, wherein the relations between the position condition (i,j) and the logic rules in the operation block range are as follows:
when i=1, j=1, the operation block range is the block position (1,1);
when i=1, j=2, the operation block range is the block positions (1,1) and (1,2);
when i=1, j=3, the operation block range is the block positions (1,1), (1,2), and (1,3);
when i=1, j≠1, 2, 3, the operation block range is the block positions (1,1), (1,2), and (1,3);
when i≠1, j=1, the operation block range is the block positions (i−1, j), (i−1, j+1), and (i−1, j+2); and
when i≠1, j≠1, 2, 3, the operation block range is selected from (a) the block positions (i−1, j−1,), (i−1, j), and (i−1, j+1) and (b) the block positions (i−1, j), (i−1, j+1), and (i−1, j+2);
an error-diffusion processing unit, which executes the error-diffusion operation for each of the pixels in the block to determine a bi-level output of the pixel for storage and to compute an average error (BE) of the block for storage; and
a storage unit, which stores the bi-level output and the average error (BE);
wherein when an unprocessed block remains in the image partition unit, the operations of the operation block processing unit, the summed pixel processing unit, and the error-diffusion processing unit on the blocks are repeated until all of the blocks are processed, and the storage unit outputs the bi-level image.
2. An error-diffusion image processing system to convert a multi-level image into a bi-level image, comprising:
an image partition unit, which divides the multi-level image into a plurality of blocks, each having a position condition (i,j) and a plurality of pixels;
an operation block processing unit, which reads in a block according to a processing order and determines an operation block range according to the position condition (i,j) to compute a summed error (BEsum) associated with the block;
a summed pixel processing unit, which determines a summed pixel range according to the position condition (i,j) of the block and adds the summed error to the summed pixel range, wherein the summed error is computed as follows:
when i=1, j=1, the summed error (BEsum) is BE(1,1)*3;
when i=1, j=2, the summed error (BEsum) is BE(1,1)+BE(1,2)*2;
when i=1, j=3, the summed error (BEsum) is BE(1,1)+BE(1,2)+BE(1,3);
when i=1, j≠1, 2, 3, the summed error (BEsum) is A*BE(1,1)+B* BE(1,2)+C*BE(1,3);
when i≠1, j=1, the summed error (BEsum) is A*BE(i−1, j)+B* BE(i−1, j+1)+C*BE(i−1, j+2); and
when i≠1, j≠1, 2, 3, the summed error (BEsum) is selected from (a) A*BE(i−1, j−1)+B*BE(i−1, j)+C*BE(i−1, j+1) and (b)A*BE(i−1, j)+B*BE(i−1, j+1)+C*BE(i−1, j+2);
where A, B, and C are weights;
an error-diffusion processing unit, which executes the error-diffusion operation for each of the pixels in the block to determine a bi-level output of the pixel for storage and to compute an average error (BE) of the block for storage; and
a storage unit, which stores the bi-level output and the average error (BE),
wherein when an unprocessed block remains in the image partition unit, the operations of the operation block processing unit, the summed pixel processing unit, and the error-diffusion processing unit on the blocks are repeated until all of the blocks are processed, and the storage unit outputs the bi-level image.
3. The improved error-diffusion image processing system of claim 2 , wherein A is 1/55, B is 1/85, and C is 1/95.
4. An error-diffusion image processing system to convert a multi-level image into a bi-level image, comprising:
an image partition unit, which divides the multi-level image into a plurality of blocks, each having a position condition (i,j) and a plurality of pixels;
an operation block processing unit, which reads in a block according to a processing order and determines an operation block range according to the position condition (i,j) to compute a summed error (BEsum) associated with the block;
a summed pixel processing unit, which determines a summed pixel range according to the position condition (i,j) of the block and adds the summed error to the summed pixel range, wherein the relations between the position condition and the logic rules of the summed pixel range are as follows:
when i=1, j=1, the summed pixel range is the last four bits;
when i=1, j=2, the summed pixel range is the first four bits and the last four bits;
when i=1, j=3, the summed pixel range is the first four bits and the last four bits;
when i=1, j≠1, 2, 3, the summed pixel range is the first four bits and the last four bits;
when i≠1, j=1, the summed pixel range is the last four bits; and
when i≠1, j≠1, 2, 3, the summed pixel range is selected from (a) the first four bits and (b) the last four bits;
an error-diffusion processing unit, which executes the error-diffusion operation for each of the pixels in the block to determine a bi-level output of the pixel for storage and to compute an average error (BE) of the block for storage; and
a storage unit, which stores the bi-level output and the average error (BE),
wherein when an unprocessed block remains in the image partition unit, the operations of the operation block processing unit, the summed pixel processing unit, and the error-diffusion processing unit on the blocks are repeated until all of the blocks are processed, and the storage unit outputs the bi-level image.
5. An error-diffusion image processing method to convert a multi-level image into a bi-level image, comprising the steps of:
(a) dividing a multi-level image into more than one block, each block having a position condition (i,j) and containing more than one pixel;
(b) reading in sequence blocks for processing and determining an operation block range according to the position condition (i,j), thus computing the summed error of the block (BEsum), wherein the relations between the position condition (i,j) and the logic rules in the operation block range are as follows:
when i=1, j=1, the operation block range is the block position (1,1);
when i=1, j=2, the operation block range is the block positions (1,1) and (1,2);
when i=1, j=3, the operation block range is the block positions (1,1), (1,2), and (1,3);
when i=1, j≠1, 2, 3,the operation block range is the block positions (1,1), (1,2), and (1,3);
when i≠1, j=1, the operation block range is the block positions (i−1, j), (i−1, j+1), and (i−1, j+2); and
when i≠1, j≠1, 2, 3,the operation block range is selected from (a) the block positions (i−1, j−1), (i−1, j), and (i−1, j+1) and (b) the block positions (i−1, j), (i−1, j+1), and (i−1, j+2);
(c) determining the summed pixel range according to the position condition (i,j) of the block, adding the summed error to the summed pixel range; and
(d) executing an error-diffusion calculation for each pixel in the block to determine the bi-level output of each pixel and storing the result, and computing the average error (BE) of the block and storing it,
wherein steps (b) and (c) are repeated until there is no unprocessed block in step (a), and the average error (BE) of the block is computed and stored.
6. An error-diffusion image processing method to convert a multi-level image into a bi-level image, comprising the steps of:
(a) dividing a multi-level image into more than one block, each block having a position condition (i,j) and containing more than one pixel;
(b) reading in sequence blocks for processing and determining an operation block range according to the position condition (i,j), thus computing the summed error of the block (BEsum), wherein the summed error is computed as follows:
when i=1, j=1, the summed error (BEsum) is BE(1,1)*3;
when i=1, j=2, the summed error (BEsum) is BE(1,1)+BE(1,2)*2;
when i=1, j=3, the summed error (BEsum) is BE(1,1)+BE(1,2)+BE(1,3);
when i=1, j≠1, 2, 3, the summed error (BEsum) is A*BE(1,1)+B* BE(1,2)+C*BE(1,3);
when i≠1, j=1, the summed error (BEsum) is A*BE(i−1, j)+B* BE(i−1, j+1)+C*BE(i−1, j+2); and
when i≠1, j≠1, 2, 3, the summed error (BEsum) is selected from (a) A*BE(i−1, j−1)+B*BE(i−1, j)+C*BE(i−1, j+1) and (b)A*BE(i−1, j)+B*BE(i−1, j+1)+C*BE(i−1, j+2);
where A, B, and C are weights;
(c) determining the summed pixel range according to the position condition (i,j) of the block, adding the summed error to the summed pixel range; and
(d) executing an error-diffusion calculation for each pixel in the block to determine the bi-level output of each pixel and storing the result, and computing the average error (BE) of the block and storing it.
7. The improved error-diffusion image processing method of claim 6 , wherein A is 1/55, B is 1/85, and C is 1/95.
8. An error-diffusion image processing method to convert a multi-level image into a bi-level image, comprising the steps of:
(a) dividing a multi-level image into more than one block, each block having a position condition (i,j) and containing more than one pixel;
(b) reading in sequence blocks for processing and determining an operation block range according to the position condition (i,j), thus computing the summed error of the block (BEsum);
(c) determining the summed pixel range according to the position condition (i,j) of the block, adding the summed error to the summed pixel range wherein the relations between the position condition and the logic rules of the summed pixel range are as follows:
when i=1, j=1, the summed pixel range is the last four bits;
when i=1, j=2, the summed pixel range is the first four bits and the last four bits;
when i=1, j=3, the summed pixel range is the first four bits and the last four bits;
when i=1, j≠1, 2, 3, the summed pixel range is the first four bits and the last four bits;
when i≠1, j=1, the summed pixel range is the last four bits; and
when i≠1, j≠1, 2, 3,the summed pixel range is selected from (a) the first four bits and (b) the last four; and
(d) executing an error-diffusion calculation for each pixel in the block to determine the bi-level output of each pixel and storing the result, and computing the average error (BE) of the block and storing it.