Algorithm for detection and recognition of traffic lights using color and contour analysis methods

Аннотация

In this paper, the problem of detection and recognition of traffic lights using color and contour analysis methods was considered and an algorithm was developed. To solve the problem, five edge detection operators are compared and used to detection and extraction traffic lights that regulate traffic for vehicles. In image preprocessing the color space of the image was converted from the RGB color space to the HSV color space. HSV color-based segmentation was performed and the resulting binary image was processed to remove noise using morphological operators.