Foreign body in order to solve the current workpiece surface morphology distribution randomness, uncertainty and high detection ability is insufficient or yan misjudgment caused by the problem, this paper respectively from the perspective of color separation characteristics and the statistical analysis, was proposed based on color separation and characteristics of the statistical analysis of surface of workpiece foreign body detection algorithm. First of all, according to brown and blue color difference, image logarithmic transformation and linear transformation, color separation equation is deduced, achieve the goal of isolated brown foreign body target. Then, based on statistical analysis principle, calculate the target color features, screen impurities interference, further accurate foreign body target detection. Finally, the algorithm was test based on the software development environment of Visual Stadium algorithm and system integration. Test results show that compared with current common foreign body detection technology, this algorithm has higher accuracy and stability.