1. Surface texture defect detection of machined parts -- detection
There are many textures on the surface of mechanical parts, and many textures are defective. If you want to find defects and compensate for defects, it is very helpful to improve the accuracy of parts. The detection in the part surface texture detection system can be carried out through the following aspects:
(1) Use some equipment and methods to test the surface principle of mechanical parts
(2) Input the detected information into the computer for processing
(3) The computer system is programmed with signal processing and spectrum image
After these three steps are completed, the routine detection of part surface texture is completed.
2. Detection and analysis of surface texture defects of machined parts
There are many time sequences in the process of machining. If there are processing defects in the time sequence, it is easy to cause defects in the surface texture of mechanical parts. However, due to different processing tools and processing methods, various textures in different forms may be produced, which not only affects the beauty but also the accuracy of parts.
After analysis, when the frequency domain filter is used to suppress the filtering operation in the region where the energy in the spectrum is relatively concentrated, the texture features in this direction will be more weak. We can reduce the texture features through this method, so as to enhance the strength of the effective texture direction, which makes it easier to distinguish and identify the background texture and the defect texture.
3. Surface texture defect detection of machined parts -- Extraction of texture features
The separation of background texture and defect texture is the focus of image recognition, and it is more convenient and effective when the image is effectively filtered. Moreover, after filtering the background image, the texture defect image will be enhanced to a certain extent. So people can better distinguish the defect texture and the background texture. However, in the process of distinguishing, it should be noted that there are many unknown situations in this case. The first is the defective target, and the second is the noise point. It is for these two reasons that necessary treatment and noise reduction should be carried out during use.
The texture information of defects is usually expressed in Fourier transform and is isolated. In the filtering processing project, the texture signal will also be automatically saved. Moreover, after image filtering, its background and texture defect characteristics are obviously different.
