Human Face Recognition Based On The Core Characteristics Under The Condition Of Illumination

A core characteristics based human face recognition method under the condition of illumination is proposed according to the problem of the sharply declining human face recognition rate under the condition of light. With this method, if human face image is affected by light and the illumination is forward or side can be judged; the images affect by illumination are processed using the strategy of frequency domain replacement, and then the key areas of human face image are divided and then are recognized using support vector machine (SVM) based on the unit of area, and finally the recognition results are integrated. The experimental result shows that this method can produce a better recognition effect than other methods in view of the problem of illumination.