JPH0530967U - Image processing device - Google Patents

Abstract

(57) [Summary] [Object] An object of the present invention is to provide a projection image processing apparatus capable of rapidly processing even a fine image having a large number of pixels. Whenever one projection amount is obtained by the projection adder circuit 13, the predetermined threshold value data 36 stored in the threshold value register 17 is compared by the comparator 18, and the magnitude relationship obtained as a result is the size flip flip block. When the line number and the magnitude result are different from the previous one stored in the CPU 19, the change point memory 15 stores the line number and the result.
It is characterized by reading to 0.

Description

[Detailed description of the device]

[0001]

[Industrial applications]

 The present invention particularly relates to an image processing device that performs projection processing.

[0002]

[Prior Art]

 In image processing, image data is added in the horizontal direction or the vertical direction, and the projection processing is often used to detect where on one coordinate axis an object exists based on the degree of change in the addition result. At this time, the point is to find the pixel address corresponding to the end part value of the object, which greatly changes the projection result. This address search basically does not cause any problem if software processing is performed in a device having a projection result stored in a memory and having a CPU that can access this memory. Most conventional devices have this kind of configuration, and the processing time is long because the projection processing period (timing) and the change point detection process period (timing) are performed at completely different timings. Tended to become. In addition, since general-purpose image processing devices do not intend to perform all the processing in real time at high speed, most of them are performed by software processing with flexibility, and from this point as well, the processing time is long. Is increasing.

[0003]

[Problems to be solved by the device]

 As described above, in the conventional method, the change point is detected from the projection result by software, so when processing a fine image with a large number of pixels, it is necessary to determine the magnitude relationship. Therefore, there was no problem in image processing aiming at high-speed processing.

Therefore, the present invention is not limited to hardware for comparison and collation, and is also capable of detecting change points in a pipeline while performing projection processing and providing an apparatus usable for real-time image processing. And

[0005]

[Means and Actions for Solving the Problems]

 In the present invention which achieves the above-mentioned object, a threshold value for comparison and collation with the projection result is set in advance, and a circuit for detecting a point where the magnitude relationship with the threshold value changes by hardware is provided. When a change is found, the image line number is stored in the memory so that the CPU can directly read the edge information of the object.

[0006]

【Example】

 Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a circuit block diagram for explaining an embodiment of the present invention. Reference numeral 10 is a CPU, 11 is a read control circuit of an image memory, 12 is an image memory, 13 is an addition circuit for projection, 14 is a memory for storing a projection result, 15 is a memory for storing a change point address, and 16 is a memory. Address generation circuit, 17 is a threshold register, 18 is a comparator (1) that compares the projection result with a reference value, 19 is an F / F (flip-flop) that remembers the magnitude relationship between the projection result and the reference value, and 20 is the magnitude relationship Of the change counter, 21 is a change point counter, 30 is a projection start command signal, 31 is a memory read address signal, and 32 is image data (usually a binarized "1" or "0" Data), 33 is an address increment signal that becomes active when reading from the image memory is completed for one line, 34 is projection data, 35 is threshold setting data, 36 is threshold data, and 37 is large. A small signal, 38 is a previous large / small signal, 39 is a change detection signal, 40 is projection line number data, 41 is change point number data, and 50 is a memory read data bus. A circuit in which data is input to the image memory is omitted, and a circuit block diagram in which only the projection processing unit is limited is shown. Also, in this figure, it is assumed that only one threshold is provided and change detection in two stages is assumed, but multiple thresholds can be set and change detection in multiple stages can be easily expanded. You can Next, the operation of the block diagram of FIG. 1 will be described.

When the projection start command signal 30 from the CPU 10 is given to the read control circuit 11, the operation of this circuit is started. The image memory 12 normally contains data such as the original image of FIG. 2 (for example, 512 × 512 pixels), and the read control circuit 11 sequentially reads these data and supplies them to the adder circuit 13 and at the same time for one line. The address increment signal 33 is output to the address generation circuit 16 each time the minute (512 pixels) is completed.

The addition circuit 13 calculates the total sum of pixel data for one line, and the projection data 34 is stored in the projection result memory 14 for each line. At the same time, it is transmitted to 18 comparators (1). Since the threshold value that divides the projection data by the CPU 10 into two is set in the threshold register 17 prior to the projection processing, the comparator (1) sets “1” when projection data 34> threshold data 36. When the projection data 34 is smaller than the threshold data 36, the output signal "0" is output, and the magnitude signal 37 is transmitted to the 20 comparator (2) and stored in the magnitude flip-flop 19. R.

In the comparator (2), the previous large / small signal 38 and the current large / small signal 37 are checked for coincidence, and when they are different, the change detection signal 39 becomes active. The number of times of activation is counted by the change point counter 21, and the change point number data 41 of this output is used as an address signal to the memory to change the line number data 40 that causes the change in the change point memory 15. And the size data 37) that is compared with the threshold value is written.

For example, in the concrete example of FIG. 2, the data obtained by projecting the hatched object of the original image is as shown by the solid line on the right side of FIG. , And j, the change point is detected twice from 1 to 0. The data written to the change point memory are i and j, and the magnitude data 1 (when i) and 0 (when J) that indicate whether the threshold value is larger or smaller after the change. It becomes like 3.

The CPU 10 can immediately know the pixel number at the end of the object by reading the change point memory 15 after completion of all the lines. Of course, since it is also possible to sequentially examine the contents of the projection result memory 14, it is naturally possible to perform a finer image analysis in combination with software processing as in the conventional method.

[0013]

[Effect of the device]

 As described above, the present invention makes it possible to detect the edge of the projection image by hardware by simply adding a simple circuit to the conventional projection processing circuit, and the processing timing is also performed simultaneously during the projection processing. Since the configuration is adopted, it is possible to provide a means for increasing the speed of image processing, and it is possible to achieve an extremely excellent effect.

[Brief description of drawings]

FIG. 1 is a circuit block diagram according to an embodiment of the present invention.

FIG. 2 is a diagram showing a specific example of projection processing according to the present invention.

FIG. 3 is a diagram showing an example of data storage in a change point memory of the present invention.

[Explanation of symbols]

10 ... CPU, 11 ... Image memory read-out control circuit, 1
2 ... Image memory, 13 ... Projection addition circuit, 14 ... Projection result memory, 15 ... Change point address memory, 16 ... Address generation circuit, 17 ... Threshold register, 18 ... Projection result as a reference Comparator (1) to compare with the value,
19 ... F / F to remember the magnitude relationship between the projection result and the reference value
(Flip-flop), 20 ... Comparator (2) for observing changes in magnitude relationship, 21 ... Change point counter.

Claims (1)

[Scope of utility model registration request] 1. In image processing for projecting, each time the projection amount of one line is obtained, it is compared with a preset threshold value, and when the magnitude relation obtained as a result is different from that of the preceding line. An image processing apparatus characterized in that the line number at that time and the magnitude result are stored in a change point memory and read into a CPU.