This study made an attempt at proposing an object edge detection method based on multi-step image enhance-ment and adaptive morphological operations to address the issues of low accuracy and unsatisfactory results in mechanical structure edge detection in low light environments.This method utilizes the multi-scale Retinex algorithm to enhance the brightness and contrast of low light images,and combines sharpening and filtering algorithms to remove noise while pre-serving details.It uses adaptive morphological operations to remove pseudo edges and strengthen weak edges.In object edge detection experiment under low light conditions,the proposed method achieves a success rate of 86.7％ which is 6 .7％ higher than conventional methods,indicating its better accuracy and quality.
维普
万方数据