Conveyor belt deviation detection method and device, computer equipment and storage medium

Abstract

The embodiment of the invention discloses a conveyor belt deviation detection method and device, computer equipment and a storage medium. The method comprises the following steps: obtaining a picture of a conveyor belt to obtain a to-be-detected picture; performing binarization semantic segmentation on the to-be-detected picture to obtain a semantic segmentation picture; performing edge straight line detection on the semantic segmentation picture to obtain a parameter equation of a straight line; determining a conveyor belt edge angular bisector; performing rolling shaft detection on the to-be-detected picture; determining the angular bisector of the reference line of the rolling shaft; determining an included angle deviation angle; determining a global offset area; determining the deviation degree of the conveyor belt; and performing alarm processing. By implementing the method provided by the embodiment of the invention, the problem that the detection cannot be performed due to the instrument deployment angle problem can be solved, the standard does not need to be artificially defined, the deviation of the conveyor belt is pre-warned, and the condition that a plurality of edges are detected or some edges cannot be detected when the edges of the conveyor belt are detected can be avoided.

Description

technical field [0001] The present invention relates to a conveyor belt, more specifically to a conveyor belt offset detection method, device, computer equipment and storage medium. Background technique [0002] Cargo conveyor belt conveyors are widely used in short-distance transportation scenarios such as shipping ports and processing plants. Its simple structure, large conveying capacity, and wide range of use scenarios make it suitable for use in all walks of life. However, during the transportation process, the cargo The transport conveyor belt will be shifted due to installation position deviation, material overweight, offset and other issues. Long-term deviation may cause the conveyor belt to break, which seriously affects the service life of the conveyor belt. Once a problem occurs, it will also affect the entire production process. [0003] The existing methods for detecting the deviation of the conveyor belt in cargo transportation are mostly contact type, laser i...

AI Technical Summary

Problems solved by technology

[0002] Cargo conveyor belt conveyors are widely used in short-distance transportation scenarios such as shipping ports and processing plants. Its simple structure, large conveying capacity, and wide range of use scenarios make it suitable for use in all walks of life. However, during the transportation process, the cargo The transport conveyor belt will be shifted due to installation position deviation, material overweight, offset and other issues. Long-term deviation may cause the conveyor belt to break, which seriously affects the service life of the conveyor belt. Once a problem occurs, it will also affect the entire production process

[0003] The existing methods for detecting the deviation of the conveyor belt in cargo transportation are mostly contact type, laser infrared, Hall magnetic induction, etc., and this type of detection method uses the card slot to limit the displacement of the conveyor belt, which will rub the edge of the conveyor belt and cause damage to the conveyor belt. Or it needs to be detected by a very expensive instrument. Once the sensor has a problem, the result of the detection system itself will be seriously affected

[0004] Chinese patent CN201911154456.2 uses image recognition to determine the first and second target rectangular areas composed of left and right rollers, and calculates their areas respectively to obtain the area ratio to determine whether the conveyor belt is offset. The advantage of this method is that no additional equipment is required. It will not cause wear and tear on the conveyor belt, but the problem is that the Hough line detection is very poor for the line detection of the defective conveyor belt, which will lead to errors in the overall accuracy

Chinese patent CN201510221251.7 is based on machine vision, marks the central axis of the conveyor belt on the non-loading surface of the cargo transport conveyor belt, and uses the angle between the vertical line of the detected conveyor belt central axis from the origin of the image coordinate system and the horizontal direction of the image coordinate system to describe the conveyor belt The deviation and direction of the deviation. The advantage of this method is that it maintains high precision without using historical sample data to fit, but it also has many disadvantages. Manually marking the central axis, some scenes are not suitable for vertical viewing angles, marking the central axis on the conveyor belt, the central axis will be blurred due to dust adsorption, or the deformation of the conveyor belt after a long period of load will cause the problem of unstable detection

[0005] To sum up, due to the complex site environment of the port in the existing technology, the existing cargo conveyor belt will wear and slip after a long period of operation, so that the final deviation will occur. At the same time, due to the limitation of the site environment, the detection The instrument cannot look directly at the conveyor belt, and the previous method only detects the conveyor belt at the angle of the front view; the conveyor belt offset detection basically uses the Hough change for straight line detection, but this kind of method often causes irregularities due to the uneven edge of the conveyor belt. For the edge detection of the conveyor belt, multiple edges are detected or some edges cannot be detected; the degree of deviation cannot be quantified by classifying the deviation, and some quantification of the degree of deviation can only measure the overall deviation It is impossible to measure the degree of local offset, which makes it impossible to give an early warning of conveyor belt offset; the detection benchmark often needs to be manually scheduled, which will make on-site deployment extremely complicated. difficult thing

Images