Research On Face Recognition Based On Array Lens

Image based face age estimation is a hot topic in computer vision and image processing.Today,especially with the constant development of an aging society,automatic face age assessment not only to the criminal investigation,image retrieval,and intelligent monitoring and other fields have a positive effect,but also care the elderly living conditions of the new strategy.Therefore,the demand for accuracy of the algorithm is constantly improved,and the improvement of hardware also brings new highlights to the research.This paper studies the face age estimation algorithm based on 3 X 3 array camera.The research contents are as follows:1.Array image preprocessing.By comparing the advantages and disadvantages of different face detection algorithms,the Haar feature and AdaBoost algorithm is used for face detection to improve the accuracy of face detection.2.In this paper,the advantages of the color camera on the face age estimation algorithm are analyzed in this paper,and the registration and fusion of the 9 array images can increase the details of the facial age features.3.This paper proposes an improved network structure of face age based on convolution neural network.Compared with the existing network structure model,the optimal model is obtained by comparing the experimental results with the fine-tuning network model.Building is suitable for the network structure,the face age estimation based on the above research contents,this article USES the array lens array of image experiment,get effective in identification speed and precision of the experiment verification,fine-tune model accuracy is 3.65%higher recognition rate.