Holder image threshold value adaption method suitable for right side and reverse side discrimination of holder

Abstract

The invention discloses a holder image threshold value adaption method suitable for the right side and reverse side discrimination of a holder. When holder judgment is carried out, the luminance of an end surface identification LED (Light Emitting Diode) lamp is not changed; an adaption threshold value method is adopted, the threshold value of an image shot by a camera is set to be changed in an allowable range, the change range of the threshold value is defined as a standard threshold value+ / -7, the standard threshold value is defined as a most proper threshold value for a holder standard component during self learning and is marked as T, and a regulation range allowed by the threshold value contains 15 values including T-7, . . . ,T, . . . , T+7; during holder image judgment, the end surface of the holder is shot, the threshold value is reduced to T, and then, the image is discriminated. The adaption threshold value method is adopted, the threshold value of the image shot by the camera is set to be changed in the allowable range, and the end surface of the same holder needs to be regulated for at most four times to realize the regulation of 15 luminance levels so as to realize image judgment at shortest time.

Description

technical field [0001] The present invention relates to the technical field of mechanical parts detection equipment, and more specifically, to a cage image threshold adaptive method suitable for judging the front and back sides of the cage. Background technique [0002] The cage partially wraps all or part of the rolling elements and moves with them to isolate the rolling elements, guide the rolling elements and keep them in the bearing. For cages on equipment with high speeds such as automotive transmissions, the requirements for dimensional accuracy are very high. If the dimensional accuracy of the cage is not up to standard during the production process, installing it on the vehicle will cause excessive wear of the cage, affecting the service life of the ball cage, or driving safety and other issues. The cage mentioned in the present invention refers to a metal member whose inner and outer surfaces are spherical and both ends are planes formed after turning. In the exist...

AI Technical Summary

Problems solved by technology

[0004] 1) Since the front and back of the cage are characterized by the size of the inner chamfer, the chamfer image must be captured by a 45-degree light source to achieve image processing; however, the chamfer shooting has higher requirements on the light source, and for some cages with relatively small chamfers , the captured image ring is very narrow, and misjudgment is easy to occur;

[0005] 2) During the production process and the automatic feeding process of the cage, there will be a lot of oil stains standing on the surface of the cage and the chamfers of the cage, which will cause the captured image to appear unclear, and even seriously Circumstances that cause the image of the ring to break, resulting in an error in image recognition;

[0006] 3) Due to the lower cost required when the equipment is used in the production process, the cage is grasped and placed under the camera by the mechanical claw composed of the cylinder and the air claw, and at the same time, a manipulator must be able to adapt to as many cages with different diameters as possible , so it is easy to cause the positioning accuracy of the cage to be low, and the angle deflection may occur after the cage is grasped, resulting in unstable reflection intensity of the light source, thereby affecting the quality of the image;

[0007] 4) Due to the processing itself, the center of the outer circle and the inner circle of the cage may not completely coincide. The method of only taking a section of the ring for identification cannot be used, and the entire image must be identified; considering the cost factor of the camera, In the actual production process, it is impossible to adopt a high-resolution camera, so it also brings difficulties to image recognition

[0008] At the same time, since the entire system has very high requirements on the reliability of image recognition, misjudgment in the production process is absolutely not allowed, and the number of suspicious products must be reduced, so this equipment cannot be well adapted to the actual production process

For this reason, the applicant devised a method to take pictures of the end faces and chamfers of the cage at the same time and use the AND operation to distinguish the front and back of the cage. The brightness of the changing time will also change. As the use time of the LED lamp increases, its temperature will change, causing the brightness of the LED lamp to change, resulting in a change in the image, so it is easy to appear in the process of image binarization. Errors lead to image feature extraction errors. For example, if the brightness of the LED is strong, the width of the image recognition ring will be larger. If the brightness of the LED is weak, other rings will be smaller or the surface of the ring will be broken. ring

The conventional method is to automatically determine the threshold value through the histogram, but this method automatically recognizes the edge of the image in actual use, but the edge it recognizes is not necessarily the edge that is actually needed, so the judgment of the edge lacks an artifact Inherent features are used as the basis, and are easily interfered by other pixels of the picture