The invention discloses a
circuit breaker quality detection
system and method based on
big data analysis, and relates to the field of
electrical equipment manufacturing and quality detection.The method comprises the steps that original images of a
circuit breaker opening and closing
handle area, an
energy storage handle area, an opening and closing indication area and an
energy storage indication area are synchronously collected through an industrial camera with a fixed
station; and a standardized detection image is generated through lens
distortion correction and adaptive brightness
equalization processing. Identifying the existence of each
nameplate by adopting a pre-trained target detection model, and judging whether a missing defect or an identification failure defect exists or not through a confidence coefficient threshold value; and for the
nameplate with qualified existence, extracting a character
direction vector by using an OCR model, performing
cosine similarity comparison with a preset standard
direction vector, and judging whether the character direction is abnormal or not. And logic or operation is performed on the two types of defect marks, if any defect exists, sound-light alarm is triggered, and structured alarm information including defect types, position coordinates and regional images is automatically generated, so that the efficiency and accuracy of quality detection of the
circuit breaker are effectively improved.