International Journal of Computer Applications (0975 – 8887) Volume 83 – No3, December 2013 30 Real-Time Traffic Sign Recognition System based on Colour Image Segmentation Vishal R. Deshmukh Research Scholar Computer Engineering Department S.S.B.T. College of Engineering And Technology Jalgoan, Maharastra, India G. K. Patnaik, Ph.D Head Of Department Computer Engineering Department S.S.B.T. College of Engineering And Technology Jalgoan, Maharastra, India M. E. Patil Associate Professor Computer Engineering Department S.S.B.T. College of Engineering And Technology Jalgoan, Maharastra, India ABSTRACT The traffic symbol system has been studies for many years. The system mainly has two phases, 1) Sign detection, 2) Recognition. It is one of the most important research area for enabling vehicle driving assistances. The automatic driving system should be simple in order to detect symbol with high responses. The challenge gets more difficult in order to make system simple while avoiding complex image processing techniques to detect symbol. The propose system consist of three main phases, 1) Frame selection, 2) Symbol detection, 3) Symbol Recognition. In frame selection phase the best frame possibly consisting of traffic symbol is selected from number of frames. In detection phase color image segmentation is perform on given frame in order detect the symbol. In recognition phase the detected traffic symbol is recognized using joint tranceform correlation (JTC) technique which is generally use for pattern matching. General Terms Vehicle, road sign, detection, color, techniques Keywords Segmentation, region of interest, detection and recognition, edge base, Joint Transform Correlation technique 1. INTRODUCTION The system is develop to first detect, and then recognize a set of Traffic sign from within frame. Such a system should be capable to studies the traffic scene image captured by the camera, obtain the traffic symbol region and make smart decisions. Generally, the system working of traffic symbol detection and recognition is divided into three stages: frame selection, segmentation and detection, recognition. Detection stage is accountable for finding region of interest (ROI) from the picture frame that is taken by a video camera mounted on top of a vehicle. ROI is the most probable part of the image that may contain traffic symbols. The second stage of automatic detection and recognition of traffic symbol. In this stage each ROI is according to their edge. In the last stage individual traffic symbol is recognized from its class. Detection algorithm can be based on template properties of traffic symbols. However, color information of input images is sensitive to the change of lighting conditions and weather. Moreover city like environment contains lots of non-symbol objects with shape and color information similar to the traffic symbols. It can increase number of symbol-candidates and increase recognition time and number of false-positive recognitions - then non-symbol object is recognized as a traffic symbol. The system will first see about those techniques that are based on finding object regions in color images. It will also see couple of techniques that deal with color segmentation to highlight how this has been used for outdoor scene analysis. 1.1 Challenges There are many problems that system has to take into account as the different conditions, blur image from moving vehicle, displacement of the traffic symbols, multiple traffic symbols blocking each other, faded traffic symbols due to effect of whether obstruction of traffic symbols by either natural or artificial elements like trees or street posters and the presence of objects having similar type of shape and color as traffic symbols. 2. RELATED WORK The region comes near attempt to single out areas of images that are uniform matching to a given set of characteristics. Given areas may be develop, shrivel, merged, split, created or destroyed during the segmentation process. There are two typical region-based segmentation algorithms: region- growing, and split-and-merge. Region growing(develop) is one of the most simple and popular algorithms and it starts by selecting a starting point or seed pixel. Then, the region develop by adding neighboring pixels that are uniform, according to a certain match criterion, increasing step by step the size of the region. Typical split-and-merge techniques consist of two basic steps. If this region does not follow with a uniform criterion the region is split into four quadrants and each quadrant is tested in the same way until every square region created in this way contains homogeneous pixels. Next, in a second step, all adjacent regions with similar attributes may be merged following other criteria. There are many options used by the system for detection and recognition of traffic symbols from the natural image. Mainly the system consist of detection phase and classification phase. The detection phase widely uses the image segmentation. When working with the gray-level images, the search is mainly based on the shape and can be quite computationally expensive. For traffic symbol detection, several methods have been developed based on shape recognition. T. Ueta and Y. Sumi and N. Yabuki and S Matsumae[10] used a self- organizing map (SOM) to extract a contour line and recognize the traffic-symbol shape from it.R. Belaroussi and J. Tarel[23] proposed a geometric model of the image gradient orientation to detect triangular symbols. R. Marmo and L. Lombardi [21] used optical flow analysis to identify the rectangular symbols and then by searching gray-level discontinuity on the image and Hough transform for detection of Milepost symbols. Loy and Barnes [22] implemented an algorithm based on fast radial symmetry, where patterns of edge orientations are exploited to detect triangular, square, and octagonal traffic symbols. However, gradient based