Addressing the challenges of limited noise suppression, inaccurate edge detection, and poor adaptability encountered in the detection of workpiece image edges in complex industrial environments, this study introduces an adaptive algorithm for edge detection that integrates morphological processing with the Canny operator. Firstly, a weighted adaptive morphological processing is designed to extract the gradient magnitude and direction. Secondly, this extracted information is weighted fused with the gradient magnitude and direction obtained from the traditional Canny operator. Subsequently, an adaptive dual threshold algorithm is devised for edge detection and connection. Finally, the proposed algorithm is experimentally validated using images collected from industrial robot assembly lines. The results demonstrate that the proposed algorithm outperforms traditional methods, detecting more effective and continuous edges with significant noise reduction.