Object detection and recognition in the driver assistace system based on the fractal analysis

The paper considers detection of artifacts in CCTV systems and recognition of the traffic signs. We propose the algorithms based on the fractal analysis to solve these problems. Proposed algorithm of artifact detection uses the fractal dimension as a shape feature of the artifact. Recognition algorithm is based on the iterative functions systems and comparison of its attractors. The efficiency of the proposed approaches was proved experimentally. Proposed algorithms are robust to the most of the image distortion such as scaling, rotation, and shift in the wide range of distortion intensity.