Study On Steel Ball Surface Defect Detector Control Technology And Image Processing Algorithms

As the main part of bearing, steel ball’s surface integrity has a decisiveinfluence on the service life of bearing, the using performance and the rotationaccuracy. In the modern equipment manufacturing industry, the bearing has avery wide range of applications. Therefore, producing good steel ball of surfaceintegrity have the practical importance, it also will make huge economic benefit,greatly reduce the consumption of resources. However in the most importantprocess of bearing steel ball production processing technology, steel ball surfacedefect detection, the current detection methods, detection efficiency andprecision are yet to be further improved. In this thesis, based on machine visiontechnology, this thesis researches on the control of the steel ball surface defectdetector and defect detection technology with following aspects:First of all, using the computer parallel port design PC control program anddesign pulse generator. The series of pulse generator has such as a single chip,PLC and motion control card, in order to reduce the development cost ofhardware and maintenance costs, design pulse generator based on PC parallelport, a variable frequency pulse sequence, drive motor to perform thecorresponding action; Design the subtle level accurate delay procedures; Receivethe external trigger control system design, improve equipment automation degree.Secondly, based on digital image processing of the surface of the steel ballimage, extract defect area. The design the eliminate noise algorithm which issuitable for the environment of different working conditions, the maximum toremove noise while preserving useful image information; Design image contrastenhancement algorithm which is suitable for the characteristics, and makes theimage background region be suppressed and the feature of steel ball details beenhanced; Design segmentation algorithm which is suitable for each image has a number of steel ball images, to separate the image of each steel ball in the targetregion from the background region; Based on a central region radiation area ofthe steel ball defect extraction method, positioned detection basis points quickly,delineating detection range.Finally, tracking and recording the overall operation of the equipmentsituation. For detection efficiency and precision, through the orthogonalexperiment analysis with testing data, choose the optimal detection parametercombination.