Traffic Sign Recognition Systems help drivers by classifying the traffic signs with some dedicated hardware. In this study, color segmentation and pattern recognition algorithms are developed for detecting speed bump traffic sign. For detection, standard featured mobile cell phone camera is used and this study is concentrated on especially high noisy situations. It is expected that algorithm is capable to choose the best threshold values for color and shape based filters. For automatic threshold decision, a histogram based analysis is made and resulted threshold values are used as upper and lower bound of Otsu Method.
In the second part of the study, the speed bump sign which is located in red triangle part of the plate is recognized by using various area and ellipticity filters. The developed algorithm can detects speed bump signs from different noisy situations successfully however algorithm cannot detect sign when the plate is blocked by a pedestrian or object. Obtained simulation results will be used as a pioneer study for the “Embedded Traffic Sign Detection System” which is planned to realize in future.