Design And Implementation Of Identifying Notes System Based On OpenCV

Identification of suspects is a measure frequently used in case handling in the public security system. After receiving suspects’ photos, investigators usually make written records of identification by means of manual composing. The facts of the case and the identification are settled through written records of identification and in the form of written reports, as so to form the evidence chain of the case. However, such approach is of low efficiency and the specifications of face photograph produced by it are not unified. In order to overcome these shortcomings and improve investigators’ efficiency in solving cases, an identification written record system was developed through facial measurement techniques on the basis of OpenCV function library in this paper. Firstly, the preprocessing techniques of images were elaborated in this paper, and then the Harr features of facial images were extracted through the method of integral images, and Adaboost algorithm was adopted to achieve the face detection based on feature extraction; and then the results of false detection were removed by the skin color model; and finally, the central points of eyes were located through eye location algorithm, and the images were cropped according to the positional relationship between the eyes and the face contour, and the results were output to the software interface. Investigators can sort the results in the image area according to the actual situation and then print out to form a written record of identification. This method is of high efficiency and strong usability, which can achieve good results in practical applications.