JPH01240294A - Image take in method for safety monitoring device - Google Patents

Abstract

PURPOSE:To shorten a processing time to one nth and heighten the reliability of monitoring of safety by conducting image processing enough for n pieces of cameras with one time image input and image processing. CONSTITUTION:As much voltage V0 as being equivalent to a brightness level 0-LEV is reduced from the image signals of cameras 1-4. Then, all other portions except the portion where brightness has been more than LEV, become less than 0V. The maximum signal is selected to be inputted into the image processing portion in a maximum value priority circuit from the image signals of 4 pieces of cameras 1-4 processed as above. This image processing portion is binary coded at the level of a brightness level >0, and its picture prime number is calculated. In this instance, in the case of respective cameras 1-4 having simultaneously caught a watched object, an image extracted at the image processing portion becomes an image added (a logic sum) with the extracted images of respective cameras 1-4. That is to say, the processing for 4 pieces of cameras can be simultaneously made with one time image input and analysis, and a processing time can be shortened to one fourth.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention provides an image of a safety monitoring device that is added to automatically operable equipment such as industrial robots and NC machine tools and that also has a visual function. Regarding the import method.

[Conventional technology]

Conventionally, this type of safety monitoring device has been disclosed in, for example, Japanese Patent Application No. 138652/1985 proposed by the present applicant,
When several cameras were installed to monitor several locations as in No. 138653, images were captured and analyzed for each camera.

[Problem to be solved by the invention]

Therefore, in the conventional example, when monitoring all locations, the processing time is equal to the image capture time times the number of cameras. Furthermore, since each image must be analyzed individually, it is necessary to process the images several times using multiple cameras, which requires a large amount of processing time.

An object of the present invention is to provide an image capturing method that can perform safety monitoring with one image input and analysis when processing images from several cameras.

[Zero means to solve the problem] Therefore, in order to achieve the above object, in the image capturing method of the safety monitoring device of the present invention, the video signal of each camera is If the brightness of the background is high, a constant voltage level that slightly exceeds the level width of the background is added.

Then, among all the video signals processed above, the maximum signal is extracted when the background brightness level is low, and the minimum signal is extracted when the background brightness level is high, and input to the image processing section. If the background brightness is low.

Binarize pixels with brightness level>0 to white (1), and when the brightness of the background is high, pixels with brightness level <max to become white (1).

The number of binarized pixels is counted to identify the safety of the monitoring area.

[For production]

In this way, what remains as an image after binarization is
This is the sum (logical OR) of all objects other than the background that exist within the field of view of each camera, that is, objects to be monitored.

Therefore, when it is only necessary to check whether there is a monitored object within the field of view of each camera, such as in a safety monitoring device,
It is possible to perform batch processing as described above, and the processing time can be reduced to 1/(number of cameras) compared to conventional analysis of each camera.

〔Example〕

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a method for capturing images of a safety monitoring device according to the present invention will be specifically explained based on embodiments shown in the drawings.

FIG. 1 shows safety monitoring in one embodiment of the present invention;
FIG. 2 is a perspective view showing a part of the configuration of the device. After setting the welding part 9 on the jig stand 8, the person retreats to a safe area and presses a welding start button (not shown), and the spot welding robot 7 operates to start welding. At this time, a safety monitoring device attached to the equipment is activated to monitor whether anyone is accidentally approaching.

The safety monitoring device is an information processing device that has at least a function of inputting a video signal, binarizing it at a predetermined binarization level, and counting white (1) pixels. Some circuitry is added before the camera input to the image processing section of this device.

As shown in Figure 1, the monitoring area 6 surrounds the dangerous area of the equipment.
A to 6d are provided, cameras 1 to 4 are installed above them,
This safety monitoring device constantly monitors whether a worker has entered the monitoring area.

In this embodiment, it is necessary to monitor four areas of the equipment, depending on the resolution of the camera, the size of the equipment, etc.
d Cameras 1 to 4 are installed in each for monitoring.

Cameras 1 to 4 are externally synchronized, operated by a synchronization signal (not shown) given from a safety monitoring device.The background is, for example, a black mat, and the histograms of the images of cameras 1 to 4 at this time are The result is as shown in Figures (a) and (b). Therefore, to check whether a worker (object to be monitored) has entered the monitoring area, the brightness level LEV slightly above the background is set as the binary level, and pixels at a level higher than that are set as white (1). , can be determined by its total number of pixels.

Therefore, as shown in FIG. 3(a), the voltage VD corresponding to the brightness levels OL and EV is subtracted from the video signals of cameras 1 to 4. Then, except for the part whose brightness was equal to or higher than LEV, all the other parts become equal to or lower than OV. From the video signals of the four cameras 1 to 4 processed in this way, the maximum signal is selected by the maximum value priority circuit and inputted to the video processing section. The video processing unit binarizes the image at a brightness level>0 and counts the number of pixels.

In this case, as shown in FIG. 5(a), each camera 1 to 4
The hand 22. is simultaneously the monitored object. Safety shoes24. When the helmet 25 is captured, the image extracted by the image processing section becomes an image 51 obtained by adding (oring) the extracted images of the cameras 1 to 4, as shown in FIG. 5(b). That is, if an image is input and analyzed once, four cameras can be processed at the same time, and the processing time is reduced to /4 compared to processing for each camera as in the past.

In addition, if a white or yellow mat is spread over the monitoring area and the background level is bright, the image shown in FIG. 2 (C).

After correcting the video signal as shown in FIG. 3(d) and FIG. 3(b), the minimum signal is extracted, and the video processing unit binarizes pixels with brightness level < m a x as white (1). good.

FIG. 4 is a control block diagram of the safety monitoring device according to this embodiment.

In the video signal generating section 40a, a video signal (raw) 401 from the camera 1 is input to a computing unit 406.

Further, the DC voltage -EO is applied to the synchronization signal 40 from the gate 403.
4 and outputs it in synchronization with the arithmetic unit 4 through the adjustment resistor 405.
Given to 06.

Here, when the background of the monitoring areas 6a to 6d is covered with a black mat, the voltage VO corresponding to the brightness level 0 to LEV is subtracted from the video signal (raw) 401, that is, a negative voltage is added (ADD) L. , and is input to the OR (maximum value priority) circuit 41 as a video signal (corrected) 402.

In addition, a video signal generating section 40b including cameras 2 to 4 is provided.
From 40d, the video signal (corrected) is similarly input to the OR circuit 41, the maximum value of these signals is given to the camera interface 42, and the video signal is input to the video mixer 47.

On the one hand, via bus 43, the CPU (Central Processing Unit)
44. Main memory 45. Signals are exchanged between the image memory 46 and the video mixer 47.
As a result of the exchange of signals, the image shown in Fig. 5(b) is displayed on the CRT48.
An image processing unit analysis image 51 as shown in FIG.

Note that when a mat with a white or yellow background is laid in the monitoring areas 6a to 6d, the calculator 406 performs subtraction (SUB) L.
, the OR circuit 41 becomes a minimum value priority circuit.

The safety monitoring device outputs an inoperable signal to the welding port bot if the total number of pixels is equal to or greater than the reference value, and outputs an operable signal if the total number of pixels is less than the reference value. The safety monitoring device performs monitoring by repeating the above operations.

Note that if the image is subjected to window processing in the processing section, the same window can be applied to each of the cameras 1 to 4. In other words, setting one window can have the same effect as setting one camera by, for example, the logical sum of four windows. Furthermore, if it is necessary to perform window processing for each camera, window processing may be performed after correcting the video signal of each camera.

〔Effect of the invention〕

As explained above, according to the present invention, image processing for n cameras can be completed with one image input and image processing, and the processing time can be reduced to 1/n of the conventional one, making it easier to monitor safety. A significant effect can be obtained in terms of reliability and cost reduction.

[Brief explanation of the drawing]

FIG. 1 is a perspective view of a safety monitoring device according to an embodiment of the present invention, FIG. 2 is a diagram showing the field of view and histograms of the input images when the brightness level of the background is low and high, and FIG. The figure is a diagram for explaining the video signal correction method according to the present invention, FIG. 4 is a control block diagram of this embodiment, and FIG. 5 is a diagram showing the state of the monitoring area and the image processing unit processed in the embodiment It is a diagram showing an image. 1 to 4...Camera 5...Safety monitoring device (image processing device) 6a to 6f...
・Monitoring area (window) 7...Welding robot 8...Jig table 21.23...Visual field 22.Hands 24.Safety shoes 31.33..Raw video signal 32.34..Corrected video signal 4' Oa~40d・Video signal generation unit 401・・Video signal (raw) 402・Video signal (corrected) 403・・Gate 404・・Synchronization signal 405・Adjustment resistor 406・・Logical sum (maximum value priority or minimum value priority)
Circuit 41...Camera interface 43...Bus 44...CPU (Central Processing Unit) 45...Main memory 46...Image memory 47...Video mixer 48...CRT 51...Image processing Part analysis image 52...Hand (top of image) 54...Safety shoes (top of image) 55...Helmet (top of image). Applicant's agent Mr. Sato

Claims (1)

[Claims] 1. A method for capturing camera images of a safety monitoring device equipped with a visual function, wherein the video signal of each camera from a plurality of cameras surrounding a monitoring area corresponds to the brightness width of the background. Signals with a constant level are subtracted or added in advance, and among the resulting output signals, if the brightness level of the background is low or high, the maximum level signal or the minimum level signal is extracted, and the brightness level of that signal is A method for capturing images for safety monitoring measures, characterized by binarizing pixels at a level greater than zero or at a level less than the maximum, counting the number of pixels, and analyzing the safety of a monitoring area.