Abstract: In this survey recognition and detection of text in traffic signs Scene structure from Canny edge detector with Hough transform and HOG descriptors is used to outline search regions inside the image, within which traffic sign candidates or samples are then found. Maximally stable extremal regions (MSERs) with HSV or HSI color thresholding are used to find a high variety of candidates, and then they're reduced by applying constraints supported by temporal and structural information. The recognition stage interprets the text contained within detected candidate regions. Individual text characters are detected as MSERs and are arranged into lines, before being taken using optical character recognition (OCR) and Support vector machine (SVM). Recognition accuracy is immensely improved through the temporal fusion of text results across consecutive frames.

Keywords: Maximally stable extremal region (MSER), Hue Saturation Value or Intensity (HSV or HSI) color, Optical Character Recognition (OCR) and support vector machine (SVM).