Aiming at the problems of low efficiency

high cost

and unstable accuracy in manual sorting of mechanical parts in traditional industry

a multi-point complex recognition and sorting system based on machine vision is designed from the perspectives of software and hardware deployment

so as to achieve the automation and intelligent upgrading of the sorting process.Following the logic of "acquisition-analysis-decision-execution"

the system collects workpiece images through an industrial camera.The images are preprocessed via grayscale conversion and filtering to enhance image quality and eliminate noise

which lays a foundation for subsequent recognition.Then

the sardine optimization algorithm-Otsu’s method(SOA-OTSU) algorithm is adopted to complete adaptive threshold segmentation.Finally

Blob connected component analysis is used to extract key information such as workpiece contour and center coordinates

realizing rapid recognition and positioning of the target. The processed workpiece information is input into an Arduino controller

which drives the robotic arm to complete the actions of grabbing

transporting

and placing.Experimental results show that the system effectively avoids human subjective errors

and significantly improves the sorting speed and accuracy

with the workpiece grabbing accuracy maintained stably at a high level.It fully verifies the feasibility and practicality of the proposed scheme

and the system demonstrates good application prospects.