JPH0781677B2 - Self-diagnosis method for safety monitoring device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention relates to a safety monitoring device having a visual function added to equipment that can automatically operate such as an industrial robot and an NC machine tool. Regarding self-diagnosis method.

[Conventional technology]

Conventionally, as a self-diagnosis method for this type of safety monitoring device,
Whether the power supply is normal or P-ROM (Programma
I was diagnosing whether the (bla ROM) check (TOTAL check) was normal or not.

However, since the image processing unit cannot be diagnosed only by these means, for example, the reference image is input, and it is confirmed by analyzing whether or not the normal feature value is calculated.

However, this method requires the time for processing the input image in addition to the time for inputting at least one image, so that the safety monitoring device checks itself while operating online. Therefore, the self-check is performed when the system is started up, that is, when the system is started, or when an automatic machine such as a robot is stopped except when performing safety monitoring.

[Problems to be solved by the invention]

As described above, according to the above method, since the robot or the like is in operation and the self-check cannot be performed during the safety monitoring, there is a problem that if an abnormality occurs, an accident will occur.

It is an object of the present invention to provide a self-diagnosis method which overcomes the drawbacks of the conventional methods and has a short processing time required for self-diagnosis of a safety monitoring device, and therefore can always check even when a robot or the like is in operation. .

[Means for solving problems]

In the present invention, first, as a normality check of the information processing section, a so-called TOTAL check for checking the P-ROM is performed and it is normal.

Next, as a check for disconnection or abnormality of the image input line from the camera, the camera image input signal is graded according to its brightness level, and the pixel distribution data representing the number of pixels at each brightness level is concentrated at the minimum or maximum brightness level. Absent.

Also, as a normality check of the image processing unit, the number of pixels at 2-3 brightness levels among the brightness levels corresponding to the background in the pixel distribution data of the camera input image is changed compared with those of the previous input image. There is something.

This is a self-diagnosis method for a safety monitoring device that is normal when all three of these conditions are met.

[Action]

By doing this, the normal check of the image processing unit including the camera can be performed only by extracting and comparing pixels of several brightness levels of the pixel distribution of the image data captured for safety monitoring. The processing time may be short, so that it can be put in the processing cycle of safety monitoring, and the normality check can always be performed.

〔Example〕

A self-diagnosis method for a safety monitoring device according to the present invention will be specifically described below based on an embodiment shown in the drawings.

As shown in the perspective view in FIG. 2, after the worker sets the welding parts 9 on the jig base 8 and moves back to a safe area and presses the welding start button (not shown), the spot welding robot 7 starts welding.

The safety monitoring device 5 has an image processing function.

The image processing function includes at least a function of introducing an image input signal from the camera 1 to the camera 4 to generate a pixel distribution, and a white after binarizing the image input signal at an arbitrary level. This is a function of counting the number of pixels of "1" level.

Then, the safety monitoring device (image processing device) 5 divides the periphery of the welding robot 7 into four regions, installs the cameras 1 to 4 in each region, and always enters the monitoring region 6. I'm watching an object.

In this embodiment, the image of the camera 1 is first captured, binarized at a lightness level slightly above the background, the brighter one is extracted as an image, and the number of pixels is counted. As shown in FIG. 3 (pixel distribution diagram), work clothes, hands, and the like are brighter than the background, so that when a worker enters the monitoring area 6, it is extracted as an image.

The safety monitoring device 5 then takes in the image of the camera 2 and performs the same processing. After that, the same process is performed by sequentially switching the camera 3, the camera 4, the camera 1, and so on. When the number of pixels of the image input by any of the cameras exceeds the set value, the operation disable signal is sent to the welding robot 7. By outputting and stopping the welding robot 7 in an emergency, it is possible to prevent an accident caused by a worker accidentally approaching the equipment.

A schematic flow in this embodiment is shown in FIG.

When the safety monitoring device 5 starts (step 10), it sets initial values such as the reference pixel number (step 11). Then, the stop flag is reset, the normal flag is set (step 12), and the image of the camera 1 is first input (step 13).
To do.

By the way, in step 13, since the parameter set such as the binarization level of the window obtained by dividing the inside of the monitoring area 6 in advance is set (step 131), the image input (step 132) is binarized at a predetermined level, and then white. The number of pixels of "1" is counted, and if the value is equal to or greater than the reference value, the stop flag is set (step 13
4) Do.

Next, it is diagnosed whether the self is normal (step 134).

The normal check flow in step 134 is shown in FIG.

In step 1343, the pixel distribution is 0 and the maximum level is, for example, 12
It is checked whether the number of pixels of 7 is equal to or more than a predetermined ratio of the total number of pixels in the field of view, and if the ratio is equal to or more than the predetermined ratio, the normal flag is reset.

Further, the procedure proceeds to step 1344, and the brightness within the background brightness level range
The number of pixels of LEV1 and LEV2 is compared with that of the previous input image, and if both the number of pixels of LEV1 and LEV2 are the same as the previous one, the normal flag is reset.

Then, it is checked whether the P-ROM check (TOTAL check) of the software of the information processing unit is normal, that is, whether the normal flag is set, and if it is in the reset state, the normal flag is reset. Is being processed by.

Then, the same process (step
14) Further, the same processing (steps 15 and 16) is performed by taking in the images of the cameras 3 and 4.

Finally, if the stop flag is set, an emergency stop signal is output to the welding robot, and if the normal flag is in the reset state, the normal signal output is turned off and the welding robot is stopped (step 17).

The safety monitoring device 5 repeats the above processing until the monitoring start signal given from the outside is turned off (step 1
8).

By doing this, the processing time required for the normality check is much shorter than the monitoring cycle processing time (image input time), so this normality check can be incorporated during the monitoring cycle, and the constant image processing unit is included. You can check the normality of the safety monitoring device.

〔The invention's effect〕

As described above, according to the present invention, the time required to check whether or not the safety monitoring device is normal is short, and therefore it is possible to always check, and the reliability of the safety monitoring device is improved. Since there is no need to perform the characteristic value analysis process for the normality check as in the past, the system can be started up and started easily, which is a remarkable effect.

[Brief description of drawings]

FIG. 1 is a diagram showing a safety monitoring general processing flow and a normal check schematic flow in one embodiment of the present invention, FIG. 2 is a perspective view of an example of a safety monitoring device, and FIG. 3 is a monitored object of the safety monitoring device. FIG. 9 is a pixel distribution diagram for explaining whether or not (worker) can be detected. 5 ... Safety monitoring device (image processing device), 6 ... Monitoring area, 7
... Welding robot, 8 ... Jig stand, 9 ... Welding parts.

Claims (1)

[Claims] 1. A self-diagnosis method for a safety monitoring device having a visual function, wherein a TOTAL check, which is a P-ROM check of an information processing unit that processes image input information, is normal, and an image input signal from a camera is received. The pixel distribution representing the number of pixels at each brightness level is not concentrated at the minimum or maximum brightness level, and the brightness level corresponding to the background in the pixel distribution of the image input signal from the camera The number of pixels at two or three brightness levels among the above is different from those of the previous input signal, and all three conditions are satisfied, and it is diagnosed as normal. Self-diagnosis method for safety monitoring equipment.