Provided is a printed circuit element polarity
machine vision detection method and a device. A guide rail used for conveying a
printed circuit board to be detected is arranged in a test area. The guide rail is provided with
stroke switches controlled by sensors. An enclosed working case covering the test area is arranged above the guide rail. The internal part of the enclosed working case is provided with a
light source capable of adjusting brightness, and multiple sensors. Different types of elements to be detected are arranged in the test area of the enclosed working case with the
light source capable of adjusting brightness in turn, and element samples are acquired.
Categorization is performed according to model numbers of capacitors or diodes, and detection parameters required by the elements to be detected are confirmed. Aiming at difference of positions of the
printed circuit board elements, the corresponding positions of different elements on the circuit board are designated, and corresponding
standard model types of different printed circuit boards are formulated via model type identification symbols, the
element model numbers, the positions and detection parameter values. A
machine vision mode is utilized to substitute a manual vision detection mode so that situations of staff negligence and errors of
leak detection of polarity directions of the polarity elements are reduced, production cost is reduced and
production quality is enhanced.