Edge Recognition Simulation of Key Features of Outdoor Dynamic Logo Image
Due to the complex lighting conditions in the outdoor environment,when the vehicle is running,motion blur will be produced in vehicle-logo images,resulting in blurred edge positioning,and affecting accurate edge recog-nition.To effectively address this issue,this paper presented a precise edge recognition method for key features of dy-namic vehicle-logo images in the outdoor environment.Firstly,we used prior knowledge to roughly locate dynamic ve-hicle logos,and used the background texture removal method to obtain precise localization images.According to RGB color space categories,we calculated the vector distance and average vector of outdoor dynamic images,thus obtaining the linear function.After calculating the categories,we analyzed the decision rules of dynamic vehicle-logo image,and carried out the segmentation and decision classification of outdoor dynamic vehicle-logo images.Moreover,we used local discrete coefficients to obtain a discrete matrix.Furthermore,we calculated the dynamic threshold and grayscale value of the vehicle-logo images.Finally,we normalized the key features of vehicle-logo image edges,thus achieving accurate recognition of the key features.Experimental results show that the proposed method can effectively improve the accuracy of vehicle-logo location,and the average recognition rate of key feature edges in vehicle-logo images is more than 90%,while effectively reducing recognition time.
NETL
NSTL
万方数据
