The Automatic Statistical Technology Research Of Passenger Flow Based On The Image Processing

As it cannot counts accurate and real-time of passenger flow at present, this paper proposes a Digital Image Recognition System combing image recognition and DSP (Digital Signal Processor) technology, which has the quick statistical speed and high recognition rate. The method is applying for the nation invention patent a instrument for passenger flow statistics based on image processing automatically.Image preprocessing and passenger flow statistics algorithm are the focus of this paper. In preprocessing aspect, a new image denoising algorithm based on wavelet transform and mathematic morphology is proposed. On the basis of in-depth study of the fuzzy enhancement algorithm, it achieves an improved histogram-based fuzzy enhancement algorithm, and it carries on the light compensation by using image normalization, the method could eliminate the adverse effects of the image recognition because of light changes. By the improved background elimination algorithm, it lays a foundation for the next step of feature extraction. In passengers'number recognition aspect, on the basis of detailed analysis of the image content, it proposes the image shape feature extraction algorithm based on the human face, which takes the characteristic point by the target sector central coordinates. It carries on the accurate passenger flow statistical based on the statistical recognition method. Experiments prove that the proposed method have obtained the better effect.According to the embedded system design criterion, design ideas and the function demand of the system, this paper carries on the reasonable module division to the entire hardware system while proposes overall project design of the system. According to the hardware components, the paper gives a detailed description of the video source processing circuit and DSP external interface circuit design principles and concrete realization ideas. On the basis of understanding software development environment and processes, it proposes some specific measures to optimize software. Compares with the present popular passenger flow automatic statistics system,the system has quick running speed because of DSP, and can achieve the real-time passenger numbers statistics. In recognition precision, with accurate statistics by using simple but effective image recognition algorithm, it has obtained a good effect with its much simpler structure, lower cost, easy to operate and convenient debugging.