Socket Connector Recognition Based on SVM with Speeded Up Robust Feature (SURF)

摘要

SURF descriptor is characterized with invariance for image shifting and changes of scale and view angle. sVM is regarded as a popular method in object recognition. In this paper, a method used in socket connector recognition based on SVM with SURF is proposed. At first, a SURF vector is constructed by extracting the invariant features. Then classification with SVM is carried out. Finally, the accuracy rate of this method is verified and validated by contrast test with SVM method with graylevel feature.