Research On Automatic Recognition Technology Of Products’ Assembly Structures Accuracy Based On Variable Step Sampling

With the improvement of requirements to products’ assembly structures accuracy, thedetection technology is improving continually. However, the assembly quality testing is still apuzzle in industry has not been well solved. Most of the existing detection systems forspecific products used fixed step method to test the objective, which caused enormous dataand failed to meet the real-time problem of on-line test.In modern industry, for some assemble complex mechanical and electrical products that areof all shapes and sizes in the surface or internal to be tested. Variable-step samplingmechanism was employed in the case of ensuring the full acquirement of effectiveinformation, inadequate data at limited orientation were used to quickly recognize structuralbody, finally the real-time problem of online test was solved. In the case of minimumrecognized size determined of every target to be tested, optimal recognized section androtating step of every objective were determined according to relativity calculation at first. Inorder to assure the information integrity of standard libraries, minimum step was chosen assampling step of qualified product from above-mentioned steps; Then, improved SIFT(scaleinvariant feature transform) algorithm was utilized to match images of products to be detectedand qualified. So as to enhance matching precision and the speed to find the optimal position,RANSAC (Random sample consensus) algorithm and binary search method were combined,the range of binary search was confirmed on the basis of refined matching points, at last, theoptimal location information of randomly-put product to be detected in standard libraries wasdetermined accurately and quickly; Eventually four representative products were detectedusing variable-step mechanism in optimal recognized section to identify the products’assembly structure is correct or not.On the premise that the accuracy rate is guaranteed, qualified products and three defective products in which exist different assembly structures are detected. Experimental results showthat, compared with fixed-step detecting method, variable-step detecting method based onoptimal recognized section can save0.017second to1.003second at average.