JPS63141186A - Bright line binarizing device for light and shade image - Google Patents

Abstract

PURPOSE:To correctly binarize an extracted bright line by determining two thresholds from density and a histogram, and binarize said line by utilizing the result. CONSTITUTION:A video camera 4 detects an image due to slit light projected vertically to an object, outputs a video signal 12 for one image and inputs this to an A/D converting circuit 7. In the A/D converting circuit 7, a video signal converted to a digital signal 13 is stored into a variable density image memory 9 and inputted to a threshold calculating circuit 8. A threshold calculating circuit 8 calculates the density histogram based on the digital signal 13 of the inputted image and determined two thresholds from the cumulative number of times frequencies. A binarizing circuit 10 uses two thresholds, removes the data for one scanning line from the light and shade image memory 9 and binarized.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a device that extracts bright lines such as slit light projected onto an object from a grayscale image and binarizes the extracted bright lines.

[Conventional technology]

FIG. 5 shows a conventional binarization device, which includes a light source 3 that projects a slit light 2 perpendicularly onto an object to be inspected 1, an image detection device 4 such as a video camera, and a detection device. It consists of a binarization circuit 5 that binarizes the image.
In the value converting circuit 5, the bright part of the input image is converted to 1'',
The dark bone is binarized as '°0''.

As the disulfide method in the above-mentioned binarization circuit 5, there are two methods: (1) determining from the peak point of brightness of the video signal, and (2) determining by performing calculations using local operators such as 3×3 pixels.

[Problem that the invention seeks to solve]

The above method (2) can be realized by directly inputting the video signal of the video camera to the binarization circuit, so it has the advantage of simple hardware. However, the conventional method employs a method of extracting bright lines from the peak brightness of the video signal for each scanning line. Therefore, when the contrast between the bright line and the background is low, the brightness peak is simply regarded as the bright line, which has the disadvantage of erroneously confirming the position of the bright line due to disturbance light (noise). Furthermore, if the bright line is too bright, smearing or blooming (overflow) may occur due to the characteristics of the video camera when bright light enters, so the brightness peak of the bright line may not be detected or may be misrecognized.

In addition, method (2) above requires A/D conversion of one screen's worth of video signals and storage in the computer's memory, and then requires a local operator to perform calculations on all pixels, so it cannot be processed in real time. There are drawbacks.

The present invention makes use of the characteristics of the method (2) above, eliminates its drawbacks, and provides a binarization device that can correctly binarize extracted bright lines even if the contrast and brightness of a grayscale image change. With the goal.

[Means for solving problems]

In order to achieve the above object, the present invention includes a light source that projects a bright line onto an object, an image detection device such as a video camera that detects the bright line as an image, and an A/D conversion device that digitizes the detected image. In the brightness wAz value conversion device for grayscale images, which has a circuit and a binarization circuit that binarizes the digitalized grayscale image data, a grayscale histogram is calculated from the grayscale image data, and two threshold values are set from the brightest one. A threshold calculation circuit is provided to determine (m, n), and
In the digitizing circuit, the grayscale image data is binarized using the threshold value (m, n) as a reference.

〔Example〕

FIG. 1 is a block diagram showing an embodiment of the present invention, in which a video camera 4, an A/D conversion circuit 7, a threshold counting circuit 8,
The grayscale image memory 9 consists of a binarization circuit 10 and a binary image memory 11.

As shown in FIG. 5, the video camera 4 detects an image by the slit light 2 projected perpendicularly onto the object 1, outputs a video signal 12 (see FIG. 2) corresponding to -images, and outputs the video signal 12 (see FIG. The signal is input to the next A/D conversion circuit 7. The A/D conversion circuit 7 digitizes the video signal 12 and outputs the digital signal 13, which is sent to the next grayscale image memory 9.
Enter and memorize.

In addition, in parallel with the above storage operation, the digital signal 13
It is manually input to the threshold planting needle calculation circuit 8. This circuit 8 calculates a density histogram based on the digital signal 13 of the input image, and determines two threshold values from the cumulative frequency. The determination of this threshold value is explained by assuming, for example, that the number of pixels in a scanning line is 256 and the gradation level of A/D conversion is 63.The theoretical length of the width of a bright line in one scanning line is 2.
For pixels, the density where the cumulative frequency from the bright side of the density histogram exceeds 2 pixels is determined as the threshold value (m) for the bright side.
Make it. Next, the darker threshold value (n) is determined to be a density that exceeds 3 to 5 times the bright line width, for example, 6 pixels (-3 times).

The binarization circuit 10 uses the above thresholds (m, n),
Retrieve data one scanning line at a time from the grayscale image memory 9 2
In binarization, when the data of the - scanning line does not exceed the darker threshold value (n), t+OIt is used. Also - the threshold value (n) for which the scan line data is darker
As shown in FIGS. 3a and 3b, the pixel position E where the density is maximum is set as IT I It, and the other positions are set as '0''.

Furthermore, - the threshold value (rn) for which the scanning line data is brighter;
When it exceeds the threshold value, as shown in Figure 4 a and b, the position of the pixel in the center of the brighter threshold is
I If, and the others are It OIf.

The binarized signal 14 binarized as described above is as follows:
It is output to the digitized image memory 11 and stored.

〔Effect of the invention〕

As described above, this invention determines two threshold values (m, n) from the density and the gray scale, and uses these to perform binarization, so it is possible to create a dark and light image with low contrast.
On the other hand, it has the effect of correctly binarizing even a dark and light image with smear or blooming. Furthermore, since it can be implemented in hardware with a simple configuration, it has the advantage of being able to perform binarization in almost real time.

[Brief explanation of the drawing]

Figure 1 is a Pronk diagram Z showing an embodiment of the present invention Figure 2 is A
The gray scale image after /D conversion, Figure a in Figure 3 is an explanatory diagram of image data of one scanning line, Figure b in the same figure is an explanatory diagram of the binarized symbol corresponding to the above image data, Figure 4 is an illustration of the image data of one scanning line. Figure a is an explanatory diagram of other image data of one scanning line, Figure b of the same figure is an explanatory diagram of the binarization symbol corresponding to the above pixel data, and Figure 5 is a configuration diagram of a conventional binarization device. be. 1...Object, 2...Slit light, 3
...Light source, 4...Video camera, 7.
...A/D conversion circuit, 8...Threshold needle calculation circuit, 9...Darkness image memory, 10...
...binarization circuit, 11...binarization image memory,
12...Video signal, 13...Digital signal.

Claims (1)

[Claims] A light source that projects a bright line onto an object; an image detection device such as a video camera that detects the bright line as an image; an A/D conversion circuit that digitizes the detected image; In a bright line binarization device for grayscale images, which has a binarization circuit for converting grayscale images, threshold calculation calculates a grayscale histogram from the grayscale image data and determines two threshold values (m, n) from the brightest one. A bright line binarization device for a grayscale image, comprising: a circuit for binarizing grayscale image data using the threshold value (m, n) as a reference in the binarization circuit.