Abstract:
Traffic sign detection and recognition (TSDR) is
one of the main area of research in autonomous vehicles and
Advanced Driving Assistance System (ADAS). In this paper a
method is proposed to detect and classify the prohibitory subset
of German Traffic Sign data set. The traffic sign detection
module utilizes adaptive color segmentation based on mean
saturation value of local neighborhood, and Circular Hough
Transform (CHT), to detect the traffic signs of input images.
After the detection stage a validation module is placed to reject
the false alarms given by the system. For the recognition stage a
deep Convolutional Neural Network (CNN) is trained to classify
the detected signs. The proposed detection and recognition
modules are evaluated on German Traffic Sign Detection
Benchmark (GTSDB) and German Traffic Sign Recognition
Benchmark (GTSRB) respectively and achieved 95.73%
detection rate and 99.93% recognition rate. Furthermore, the
proposed system is immune to change of illumination and is able
to detect damaged, occluded and distorted traffic signs under
complex traffic environment.