Detection Method of Smart Meter Carrier Module Pin Based on Blob Analysis

Abstract

In order to achieve automatic detection of pin defects in smart meter carrier module, a machine vision detection method based on Blob analysis is proposed in this paper. Firstly, the interface region of carrier module is segmented using the maximum class variance method combined with perspective transformation. Secondly, a two-stage identification strategy based on Blob analysis is proposed to complete the rough positioning of the pins, which takes the distribution rules of the pins as constraint conditions. The spot is extracted by adaptive binarization strategy. Finally, the centroid method was used to calculate the spot center of the pin, which is registered with the template point set. The defect detection of the pin was completed according to the offset of each pin.