JPS61180382A - Picture recognizer - Google Patents

Abstract

PURPOSE:To recognize the halftone pictures like blur of a character by using a circuit which digitizes the luminance level of a video signal to integrate the output of the digitization and decides the picture luminance from the value of said integration. CONSTITUTION:The TV signal V is binary coded and stored in a frame memory 15. While the video signal is binary coded and at the same time converted into the digital quantity by an A/D converter 7. Then the luminance level of the video signal is integrated by an ALU 9. Based on this integrated value, a CPU 14 decides whether the value is larger than a fixed level or not. Thus the luminance within an area is decided for recognition of the character blurs, the background, etc. Thus it is possible to decide accurately the propriety of the external appearance of an object to be measured. This can improve extremely the recognizing performance.

Description

[Detailed Description of the Invention] , Part 1! ■D11! The present invention relates to an image recognition device that recognizes the pattern of an object to be measured using a binary image and also determines the brightness of the object.

For example, in order to determine only the presence or absence of the object to be measured, when recognizing the pattern of the object to be measured, the video signal of the object to be measured is binarized. To pattern recognize the character A, set the region (C1) of the character A to "I", the background region (C2) to 10", and set the video signal to 2".
Convert it into a value and determine its existence in characters. The circuit (1) shown in FIG. 3 is used for the binarization processing. That is, in the circuit (1), a synchronizing signal is first extracted from the television signal (V) by the horizontal and vertical synchronizing separators (2) and (3), and the video signal is input to the amplifier (4).

Then, the brightness level of the video signal amplified by the amplifier (4) is compared with a predetermined threshold level (V TII) by the comparator (5), and the video signal is
Value. When this binarized signal is combined with a synchronization signal in a video signal generation section (6), a binarized television signal is obtained.

By the way, the brightness of the image obtained by the above-mentioned binarization process is only in the '1' or '0' region, and the brightness of the image is only in the '1' or '0' region, and the brightness of the image is only in the '1' or '0' region. The brightness level is also “l” or “
Therefore, as shown in Figure 2, even if part of the letter A is blurred (C3), if it is processed as "1" through binarization, it will not be determined as blurred after processing. Also, if it is processed as "0", it will be incorrectly recognized that part of the letter A is missing after processing, and it will also be impossible to judge whether it is blurred.Therefore, conventional binarized image processing With this method, it is not possible to judge the luminance level of the intermediate tone of the appearance of the object to be measured.

The present invention provides an A/D converter for converting an analog video signal into a digital value, and an A/D converter for converting an analog video signal into a digital value, in a circuit that binarizes a video signal when pattern recognition is performed on a measured object using a binary image. An integrating unit that integrates the brightness level of the video signal digitized by the A/D converter for those included in a predetermined area of the screen; a control unit for an integrating unit that controls the integrating unit to integrate the luminance level of the integrating unit;
The apparatus further includes a processing section that determines the brightness of the object to be measured based on the integrated value in the integration section.

The brightness level of the video signal of the object to be measured, which has been digitized, is integrated, and the brightness of the object to be measured is determined from the integrated value.

One embodiment of the present invention will be described below with reference to FIG. The same reference numerals as in FIG. 3 indicate the same parts, and the explanation thereof will be omitted. In FIG. is a video signal generation unit, (7) is an A/D converter that converts the video signal extracted from the television signal into a digital quantity, (8) is a latch that holds the output signal of the A/D converter (7), ( 9) is an ALU that integrates the brightness level of the digitized video signal, and (10) is an ALU (9) that integrates the brightness level of the digitized video signal by sequentially repeating it in ALU (9). A latch that holds the output, the integration unit (18) consists of a latch (8) (10) and an ALU (9). (11) is a synchronization unit separated in the horizontal and vertical synchronization separation units (2) and (3) A screen coordinate counting section measures screen coordinates by counting signals and don't clock signals (P) corresponding to pixels, and (12) sets a specific area of the screen from the screen coordinates measured by the screen coordinate counting section. The screen area select section (13) is an ALU control section that controls the ALU (9) to integrate the brightness level of the video signal within the area set by the screen area select section (12). The control unit (17) consists of a screen coordinate counting unit (11), a screen area selection unit (12), and an ALU control unit (13). (14) integrates the video signal brightness level of a specific area in the ALU (9). The CPU (14) is a CPU that determines the brightness of the image included in the area based on the value and controls the screen area selection section (12) and the ALU control section (13).

(15) is a frame memory for storing the binarized video signal in order for the CPU (14) to judge the brightness in a specific area of the binarized video signal by the comparator (5); 16) is a D/A converter that converts the threshold level (VTH) sent from the CPU (14) into an analog quantity and inputs it to the comparator (5).

The operation of the present invention will be described below based on the above configuration.

First, the synchronization signal is separated from the television signal (V) by the horizontal and vertical synchronization separators (2) and (3), and the amplifier (
4) The video signal amplified by the comparator (5)
After binarizing the video signal, it is stored in the frame memory (15).The video signal is then binarized and converted into a digital quantity by the A/D converter (7), and the brightness level is sent to the ALU (9). and calculate the total as follows. That is, A.L.
Since U (9) adds two numbers, let the two input values be ■, ■, and the added output value be O, and then A.
/D converter (7) output is input. First, A/D
One luminance level of the video signal from the output of the converter (7) is inputted to ■, and O is inputted to ■, and the added output value @ is held in the latch (10). Then, the brightness level of the next video signal output from the A/D converter (7) is held in the latch (8) using a sampling clock, etc., and from there, this brightness level is input to ■ and the timing is determined. and input ■ held in the latch (10) to ■. By repeating the addition of the luminance levels of the digitized video signals in this manner, the luminance levels are integrated.

The luminance level integration by the ALU (9) is performed for a specific area of a part of the screen. That is, by counting the horizontal and vertical synchronizing signals and the dot clock signal (P) corresponding to each pixel using the screen coordinate counting section (11), the coordinates of the screen can be measured. Therefore, using the above screen coordinates, the screen area select section (1
In 2), by setting a specific area on the screen and inputting its output signal to the ALU control unit (13), the ALU (9) and the latch (8) (10) are used to integrate the brightness level within the area. control.

Using the integrated value, the CPU (14) determines, for example, whether the value is above a certain value, thereby determining the brightness within the area and recognizing blurred characters, backgrounds, etc. At this time, in order for the CPU (14) to make a judgment, it is necessary to store the image, so the video signal corresponding to each pixel is loaded into the frame memory (15), and the brightness of a specific area is calculated from that memory. to decide.

Then, the binarized video signal and the synchronization signal are combined in a video signal generation section (6) and a binarized television signal is output.

According to the present invention, in a circuit that binarizes a video signal when pattern recognizing an object to be measured using a binarized image,
Since we have installed a circuit that digitizes the brightness level of the video signal, integrates the output values, and then determines the brightness of the image from the integrated value, it is possible to detect blurred characters, scratch marks on semiconductor chips, etc. that cannot be detected using binary image processing. It becomes possible to recognize difficult half-tone images, and it is possible to accurately judge whether the appearance of the object to be measured is good or bad, and the recognition performance is dramatically improved. In addition, since it is a real-time process, high-speed processing is possible, and cost performance is also very high.

[Brief Description of the Drawings] Fig. 1 is a block diagram of an embodiment of the image recognition device according to the present invention, Fig. 2 is a plan view of an example of an object to be measured displayed on a monitor screen, and Fig. 3 is FIG. 2 is a block diagram of a disulfide treatment circuit. (1) - Binarization circuit, (7) - A/D converter, (1B) - Integration unit, (17) - Integration unit control unit, (14) - Processing unit. Figure 2 Figure 3

Claims (1)

[Claims] (1) In a circuit that binarizes a video signal when recognizing a pattern of a measured object using a binarized image, an A/D converter that converts an analog video signal into a digital quantity; an integrating unit that integrates the brightness levels of digitized video signals included in a predetermined area of the screen; An image recognition device comprising: a control section for an integration section that controls the integration section; and a processing section that determines the brightness of an object to be measured based on the integrated value in the integration section.