JPS63122381A - Picture processor - Google Patents

Abstract

PURPOSE:To obtain stable binarized picture data in a real time even when the quantity of light of an object is changed by sequentially adding with the weight of one picture element to a screen to which digital picture data is set by an average circuit to obtain a threshold. CONSTITUTION:The digital picture data D1 is inputted to an average circuit 3, sequentially added with the weight of one picture element to one screen to obtain average value data. Thereafter, digital threshold data D3 obtained by adding offset data D5 from an offset adjusting circuit 5 to the average value data obtaining in the average circuit 3 is outputted to a D/A converting circuit 6 to obtain threshold data D6. A comparison circuit 7 analog picture D2 according to a raster scanning from a camera 2 with said analog threshold data D6 to output the binarized picture data D7. Thereby, a threshold at the time of binarizing has a linear relation to the quantity of light of the picture and the stable binarized picture data can be obtained in a real time even when the quantity of light of the object is changed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to an image processing device that binarizes image data in order to separate an object and a background from a grayscale image.

(Prior Art) In this type of image processing device, as an example of a method for determining a threshold value when binarizing image data from a camera, a light receiving element made of CdS or the like that detects light changes in an image is used. A so-called hardware processing method is provided to obtain a threshold value based on the light-receiving voltage signal from the light-receiving element, or image data from the camera is temporarily stored in a frame memory, and based on this image data, CPU as control circuit
1: So-called software processing is provided in which an average brightness value is calculated and used as a threshold value, or a histogram is created and a threshold value is determined from this.

(Problem to be Solved by the Invention) However, in the former hardware processing method, since the light receiving voltage signal from the light receiving element does not directly reflect the amount of light incident from the camera, the threshold value is determined based on the light receiving voltage signal. The value does not have a linear relationship with the light beam for one screen from the camera, and if the contrast of the object is small and the amount of light for one screen changes, it is difficult to obtain stable binary image data. The disadvantage is that it cannot be obtained. In addition, the latter software processing method requires time to store the image data in the frame memory and then reads out the stored data to determine the threshold value, which has the problem of lacking real-time responsiveness. .

Therefore, the L1 feature of the present invention is that the threshold value for binarization has a linear relationship with the light amount of the image, and stable binary image data can be generated in real time even if the light amount of the object changes. The purpose of the present invention is to provide an image processing device that can be used to obtain images.

[Main features of the invention (Means for solving the problems) The image processing device of the present invention is an A/D conversion device that converts analog image data output from a camera by raster scanning into multi-tone digital image data. In addition to providing a circuit, an averaging circuit is provided that sequentially adds the digital image data with a weight of one pixel for a set screen and outputs a threshold value, and further adds analog image data by the next raster scan from the camera. A comparison circuit is provided which compares the threshold data derived from the averaging circuit and D/A converted and outputs binary image data, and sequentially outputs the binary image data output from the comparison circuit. It is characterized by the provision of a memory for storing data.

(Operation) A threshold value is obtained by sequentially adding digital image data using an averaging circuit with a weight of one pixel for a set screen.

Therefore, since this threshold value has a linear relationship with the amount of light for one screen from the camera, stable binary image data can be obtained. Further, since the averaging circuit only sequentially adds digital image data to +11, unlike software processing, the time for calculating the threshold value can be shortened and binarization can be performed in real time.

(Example) An example of the present invention will be described below with reference to the drawings.

In FIG. 1, 1 is an A/D conversion circuit, which manually converts analog image data D2 outputted by raster scanning from camera 2 into A/D to generate multi-gradation digital image data DI. It is designed to be output. Reference numeral 3 denotes an averaging circuit to which the digital image data DI is applied via the pass line 4, and this data D1 is applied to a set screen, for example, with the weight of one pixel for one screen, for example, one screen. When is 256×256 dots, the dots are sequentially added with a weight of 1/65,536. For example, if one screen is 256 x 256 dots as described above, the averaging circuit 3 is configured with, for example, a 24-bit addition circuit, and the averaging circuit 3 sequentially processes the data for one screen by using the digital image data D1 as 8-bit data per pixel. When added, the data for the upper 8 bits of the averaging circuit 3 becomes the average value data of the pixels for one screen.

Further, 5 is an offset adjustment circuit which outputs offset data D consisting of a digital signal indicating a predetermined offset value and supplies it to the averaging circuit 3.

That is, when the optimal threshold value is higher than the average value data obtained by the averaging circuit 3, such as when the brightness and area of the object to be recognized are brighter and smaller than the background, respectively, the difference between the two values is Since it remains constant even if the brightness of the entire screen changes, a predetermined off-set value is set to obtain the optimal threshold value. - Offset data DS indicating the value is given to the averaging circuit 3 and added to the average value data (or subtracted in the opposite case). Therefore, the averaging circuit 3 outputs data obtained by adding the offset data D5 to the obtained average value data as digital threshold data D3. 6
is a D/A conversion circuit which receives the digital threshold data D3 from the averaging circuit 3, converts it into a D/A, and outputs it as analog threshold data D8.

Reference numeral 7 denotes a comparison circuit, which manually inputs the analog image data D2 obtained by the next raster scan from the camera 2 and the analog threshold data D6 from the D/A conversion circuit 6, compares them, and calculates, for example, D2. When ≧D6, rlJ, D2
When <Da, binary image data D7 consisting of "0" is output. Reference numeral 8 denotes a memory, for example a frame memory, which manually receives the binarized image data D7 from the comparator circuit 7 via the pass line 4 and sequentially stores it. 9 is a control circuit, for example C
PU, which transmits the binarized image data Dτ stored in the frame memory 8 to the CR, which is a display device, via a pass line 4.
The information is applied to the T display 10 for display. Further, the CPU 9 is configured to perform various controls such as the timing of starting and ending the addition in the averaging circuit 3, and the timing of storing the henivalued image data D from the comparison circuit 7 into the frame memory 8.

Note that an operation mode is also set between the camera 2 and the CRT display 10 in which the analog image data D2 from the camera 2 is directly applied to the CRT display 10 via the data line 11 and displayed without being subjected to binarization processing. ing.

Next, the operation of the above configuration will be described.

Analog image data D2 outputted from the camera 2 by raster scanning is converted into multi-gradation digital image data D1 by an A/D conversion circuit 1.

This digital image data D1 is input to the averaging circuit 3 and is sequentially added with a weight of one pixel for one screen (with a weight of 1/65.536) to obtain average value data. Thereafter, digital threshold data D3 is obtained by adding offset data D5 indicating a predetermined offset value from the offset adjustment circuit 5 to the weekday value data obtained by the averaging circuit 3.
The signal is outputted to the /A conversion circuit 6 and subjected to D/A conversion to obtain analog threshold data D6. Here, the CPU 9 operates the average circuit 3 so that the D/A conversion is performed within the VD retrace time.
control the output of Then, the comparison circuit 7 compares the analog image data D2 obtained by the next raster scan from the camera 2 with the analog threshold data DB, and outputs binary image data D7. 20 After the binarized image data D7 is sequentially stored in the frame memory 8, the binarized image data D7 stored in the frame memory 8 is displayed on the CRT display 10 by the CPU 9.

The measurement results of the above example are shown in FIG.

That is, FIG. 2(a) is a luminance value distribution histogram of each pixel when the entire image is dark, and the dashed-dotted line in the figure indicates the threshold value (average luminance value+offset value). Also,
Figure 2(b) is a luminance value distribution histogram of each pixel when the entire image is bright, and the dashed-dot line in the figure is the threshold value (
average brightness value + offset value). Therefore, as is clear from FIGS. 2(a) and 2(b), the threshold value according to this embodiment is the optimum value for binarizing the background and the object.

According to this embodiment having such a configuration, the averaging circuit 3 sequentially adds the analog image data D2 with the weight of one pixel for one screen to obtain the digital threshold data D3, which is different from the conventional method. Unlike hardware processing, the digital threshold data D3 has a linear relationship with the brightness of the image, making it possible to obtain stable binary image data even for objects whose light intensity changes. I can do it.

Moreover, since the digital threshold data D3 is obtained from the averaging circuit 3, unlike conventional software processing, binarization can be performed in real time (VD and VD at the same period).

As shown in Figure 2 (C), when the contrast is low, that is, when the amount of light on the background and the object is similar, the difference between the area of the background (number of pixels) and the area of the object is If there is a large difference in the values, it is difficult to binarize. In such a case, the screen is divided into two or four, and an averaging circuit like the one in the above embodiment is provided for each divided screen, and a binary value is calculated at each threshold value for each divided screen. By performing the conversion, a good binarized image can be obtained.

[Effects of the Invention] As is clear from the above description, the present invention includes an averaging circuit that sequentially adds digital image data with a weight of one pixel for a set screen, and also adds an averaging circuit that sequentially adds digital image data with a weight of one pixel for a set screen, and A comparison circuit is provided that compares the image data with the D/A-converted threshold data derived from the averaging circuit and outputs binary image data, and the binary image data output from the comparison circuit is provided. Since the configuration is such that the values are sequentially stored in the memory, the threshold value for binarization has a linear relationship with the light intensity of the image, and stable binarization is possible even if the light intensity of the object changes. This has the excellent effect of being able to obtain image data in real time.

[Brief explanation of the drawing]

FIG. 1 is an overall block diagram showing one embodiment of the present invention, and FIGS. 2(a) and (b) are characteristic diagrams showing luminance value distribution histograms in different states of the same embodiment, respectively.
Figure (c) is a characteristic diagram showing a brightness value distribution histogram in another embodiment of the present invention. In the drawing, 1 is an A/D conversion circuit, 2 is a camera, 3 is an average circuit, 7 is a comparison circuit, 8 is a frame memory (memory), 9
10 indicates a CPU (control circuit), and 10 indicates a CRT display (display device). Figure 1

Claims (1)

[Claims] 1. A/A that converts analog image data output by raster scanning from a camera into multi-tone digital image data.
a D conversion circuit; an averaging circuit that sequentially adds the digital image data with a weight of one pixel for a set screen portion and outputs threshold data; A comparison circuit that outputs binary image data by comparing the D/A-converted threshold data derived from the averaging circuit, and a memory that sequentially stores the binary image data output from this comparison circuit. An image processing device comprising: