JPS5913786B2 - Image signal encoding processing method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an image signal encoding processing method, particularly to an image signal encoding processing method for encoding line figures.

In the conventional line figure encoding processing method, a line figure is approximated by a straight line and a circular arc, and in the case of a straight line, the coordinates of the starting point and ending point are
In the case of a circular arc, there are two methods: a method in which the coordinates of the start point and end point and the radius (or center coordinates) are converted into commands, a one-line figure is traced pixel by pixel, and the direction of movement is sequentially encoded (chain encoding method, etc.). There is.

The former method is inappropriate for complex and detailed (line figures), and the latter chain encoding method seems to be more suitable.However, in the conventional chain encoding method, line figures are When tracing and encoding a pixel string, each stroke was encoded using 3 bits (8 directions) assuming that the tracing direction was always isotropic.However, in reality, line figures Since there is a statistical property that there are many cases where progress continues in the same direction, there is a possibility that the amount of codes can be further reduced by utilizing this property.The present invention aims to improve the coding efficiency of the chain coding method. The purpose of this method is to perform chain encoding using the statistical properties of line figures, and will be explained in detail below with reference to the drawings.

FIG. 1 shows the premise of image signal encoding according to the present invention.
This shows the numbering of pixel columns.

The direction code ``O'' is given when there is a pixel adjacent to the pixel shown in the figure in the horizontal right direction, and the direction code ``1'' is given when there is a pixel in contact diagonally upward to the right. Then, direction code “O” and “IL”1
” and “2”, ..., “7” and “O”, the eight combinations that can be made by combining adjacent directions are numbered as one threat 1 as shown in Figure 2. Let's call the numbers octant numbers. Generally, the tracking directions of pixel rows that make up a line figure are "O" and "1".
For example, adjacent directions such as "7" and "O" are "O", "
It is often repeated like ``1'', ``O'', and ``1''. In the case of this repetition (V), it is possible to divide a pixel column by using the same octant number and a pixel column in which directions belonging to the same octant number are continuous as one segment. Therefore, the pixel row is divided into such segments, and the tracking direction for each pixel is set to two directions belonging to the octant, for example, "O".
” direction or the “1” direction is encoded with 1 bit. FIG. 3 shows an example image signal code string obtained by the present invention.

1 is the starting point coordinate of the line figure, 2-1, 2'-1, 7-1, ・
...indicates the octant number of each segment, 2
-2, 22-2, 7-2, . . . are the chain lengths, 2-3, 25-3, 7-3, . . . are the direction bits of each pixel, and 3 is the direction bit of each pixel. Indicates the end code of the line shape.

Assume that the chain length of a segment is represented by, for example, 8 bits, and that the average chain length of a segment is n pixels.
Since the octant number is represented by 3 bits and the direction data is represented by 1 bit, the simple chain encoding method requires a code amount of V) 3 bits per pixel, whereas in this encoding method, (3+8+n)/n= 1+(11/n) bits are required. In the case of n>>1, the amount of code required is approximately 1 bit per pixel. FIG. 4 is a specific example of tracing a line-shaped pixel column, in which the segment octant number is 01 and the chain length is 6 from the pixel shown in the figure a to the pixel shown in the figure b.

Also, the direction 0 of the 0th octane is bit 50',
If direction 1 is represented by bit number 2, the direction bit string becomes 011101 in this case. FIG. 5 shows the configuration of an embodiment of the present invention, in which 4 is an image input device, 5 is an image storage device such as a frame memory, 6 is a start point/end point detector, 7 is a start/end point encoder,
8 is a direction detector, 9 is a direction encoder, 10 is a direction code separator, 11 and 12 are first and second direction registers, respectively, 13 is an octant encoder, 14 is a direction bit encoder, 15 is a counter, 16 is a comparator, 1r is a register,
18 represents a chain length comparator, and 19 to 22 represent code string sending terminals. To explain this operation, a line figure inputted from the image processing device 4 and stored in the frame memory 5 has a starting point a (FIG. 4) detected by the starting point/end point detector 6. is,
The start point coordinates are encoded by the 8-point/end-point encoder 7 and sent out from the start-end point code string sending terminal 22. Next, in the line figure stored in the frame memory 5, the direction 011101 of the pixel column is sequentially detected by the direction detector 8, and encoded by the direction encoder 9. The first two of the obtained direction code strings are stored by the direction code separator 10 in the first direction register 11 and the second direction register 12, respectively. The octant encoder 13 reads the contents of the first direction register 11 and the contents of the second direction register 12.
As shown in the figure and FIG. 2, it is encoded to which octant direction the pixel column being processed belongs. Furthermore, the direction bit encoder 14 encodes which direction within the octant these first two directions are, each with one bit. Furthermore, the direction bit encoder 14 operates on a counter 15 so that each time the direction bit increases by one, the counter 15 is incremented.
Increment the contents by 1 and store the chain length of the segment. The third and subsequent direction codes are determined by the comparator 16.
and the contents of the second direction registers 11 and 12.

If this direction matches the contents of either of the two direction registers, the direction bit encoder 14 selects the first direction and the second direction.
Using the same code used for direction discrimination, encoding is performed with one bit per direction and stored in the register 17. At this time, the value of the counter 15 is incremented by 1 as explained in section 7. If the direction code obtained by the comparator 16 does not match any of the two direction registers 11 and 12, the comparator 16 sends a flag signal to the octant encoder 13, and the octant code is sent to the octant code string sending terminal. Send from 19. At the same time, the comparator 16 is connected to the chain length comparator 18.
A flag signal is also sent to the counter 15, the contents of the counter 15 are read, the chain length is encoded, and the chain length code string sending terminal 21
Send it out from here. Furthermore, the comparator 16 sends a flag signal to the register 1r in which the direction bit code string is stored, and sends out the stored direction bit code string from the direction bit code string sending terminal 20. When these series of operations are completed, the comparator 16 stores the direction code obtained at the end that does not belong to the previous octant in the first direction register as the first direction of the new octant, and performs the above operation. Repeat the action? vinegar. When the end point of the line figure stored in the frame memory 5 is detected by the start/end point detector 6, the processing of the final segment is finished, and the octant code, chain length,
After sending out the direction bit code string, the start/end point encoder r sends out from the start/end point code string sending terminal 22 a code that indicates the end of the line figure, not the coordinates of the end point of the line figure, and can be distinguished from a segment code. . In addition, in the process of generating an octant, if the contents of the first direction register 11 and the second direction register 12 are direction codes that do not belong to the same octant, only the direction code in the first direction register is used to generate one direction code. Configure segments.

In this case, the chain length is 1. Then, the contents of the second direction register are transferred to the first direction register, the next direction code obtained by the direction encoder 9 is stored in the second direction register, and the operation is started. Also, the first direction register 1
If the contents of the first direction register 1 and the second direction register 12 are the same direction code, when a different direction code is obtained, the direction code is stored in the second direction register and octant encoded by the octant encoder 13. do. At this time, the direction bit codes and chain lengths are integrated even before a different direction boat number is obtained. As described above, according to the present invention, an image signal can be encoded with a small amount of code, so that the storage capacity of a memory for storing image signals can be small.

Another advantage is that the transmission capacity of the transmission line is small when transmitting image signals.

[Brief explanation of the drawing]

FIG. 1 is an explanatory diagram for explaining the numbering of pixel columns used in one embodiment of the present invention, FIG. 2 is an explanatory diagram of the correspondence between direction codes and octant numbers used in the same embodiment, and FIG. FIG. 4 shows a specific example of line-shaped pixel string tracking, and FIG. 5 shows a configuration diagram of an embodiment of the invention. 1... Starting point coordinates of the line figure, 2-1, 2 and 1, 7-
1, ......: Octant number code of each segment, 2-2, 2 to 2, 7-2, ...... Chain length code of each segment, 2-3, 2'-3, Z-3c...
...: Direction bit of each pixel, 3: End code of line figure,
4: Image input device, 5: Frame memory, 6: Start point end point detector, r: Start point end point encoder, 8: Direction detector, 9: Direction encoder, 10: Direction code separator, 11: First direction register , 12: Second direction register, 13: Octant encoder, 14: Direction bit permitter, 15: Counter, 16:
Comparator, 1r: Register, 18: Chain length comparator, 1
9: Octant code string sending terminal, 20: Direction bit code string sending terminal, 21: Chain length code string sending terminal, 22:
Start point end point code string transmission terminal.

Claims (1)

[Claims] 1. In an image signal encoding processing method that encodes a given line figure, an image input device that inputs the line figure;
It has a frame memory for storing line figures input from the image input device, and a means for tracing the line figures accumulated in the frame memory as a pixel column, and has means for tracing the line figures stored in the frame memory as a pixel column, and has a frame memory that stores line figures input from the image input device, and means for tracing the line figures stored in the frame memory as a pixel column. An image characterized by forming a set of pixel columns in a direction, and encoding a line figure in a chain by using the direction of the pixel column, the pixel connection length, and the variation in direction within the direction. Signal encoding processing method.