Face Detection And Location Algorithm Based On DSP

Human face detection is to locate the faces and determine their sizes and numbers in an image. Human face detection problem, originated in face recognition as its first step to locate faces in the image, has recently been intensively researched as an independent problem because of its applications in many fields including security access control, smart visual surveillance, perceptual interface, content-based information retrieval, advanced human and computer interaction.According to face detection and location in the smart video surveillan -ce system. Algorithms and DSP implement of face detection and location are studied. The paper's major works are as follows.1. For the problem to achieve face detection and location algorithm, this paper proposed a skin detection algorithm based on the YUV color space. We have established the YUV skin color model, through face skin color's cluster analysis in the YUV color space. The simulation results show that the model can be detected in the color image region, the human face rough positioning laid the foundation.2. In view of the problem the skin color is not necessarily a person face in image, we first determinate candidate face regions using skin color model, then achieve face region by distinguishing and merging these candidate regions based on some rules. For the single face situation, projection method is used to find the regional centre and then locate the face position. For the uncertain circumstances, this paper presents multi-human face location based on the improved connected components labeling algorithm. simulation results show effectiveness of the algorithm.3. This dissertation realizes the above algorithms on DSP image processing system. algorithms are studied in MATLAB, and then are transplanted on the DSPs by Using effective image storage and algorithm structure suit DSP processing. The actual test results indicated that the algorithms are correct and feasible on the DSP.The realization of the face detection and location algorithm based on DSP has the practical application value in the smart visual surveillance system.