Millimeter Wave Image Detection Method Based On Convolution Neural Network

Artificial neural network is simulated scientists simulate artificial intelligence approach to building or part of the brain function of system simulation. In the 20 th century the technology has made great progress and then gradually decline. The discipline comes from bionics and simulation animal neurology, through systems and artificial intelligence to mimic the operation mode of human or animal vision and other brain functions, like this artificial nervous system can be viewed as an animal brain systems thinking An analog. A new artificial neural network convolution neural network artificial neural network and deep learning technology,production, and local sensing area, level, feature extraction and classification process of combining global training and other characteristics, widely used in the field of image recognition get. Convolution neural network development after decades, it has been a special part of the depth of the network widespread concern of scholars, scientists hope to present an image recognition system to be classified according to different data recognition and processing tasks they are responsible,along with the gradual deepening of convolution neural network research in even more practical significance in the field of simulation based on artificial neural networks much attention.Based on previous exploration and research on artificial neural networks to organize, focus on the significance of convolution neural network concepts and calculation methods are summarized. Meanwhile basic theory of classic swing convolution on artificial neural networks, focuses on the task of establishing the neural network model by handwritten numeral recognition and facial recognition,mainly as follows:1, the first neural network to sort out the scientific development, analysis shows the advantages and disadvantages of convolution neural network analysis pushed to its formula and operation principle.2, followed by convolution neural network handwriting database training,analysis and comparison of the results, conclusions.3, again convolution neural network to classify the data set millimeter wave body training, and at random picture identification mark items position.