Efficient high-resolution defect nondestructive testing method based on convolutional neural network

The invention relates to an efficient high-resolution defect nondestructive testing method based on a convolutional neural network. The method comprises the following steps: transmitting a plane wave with a deflection angle of 0 to a measured workpiece through an ultrasonic phased array, collecting scattered echo data of the transmitted plane wave, performing time domain filtering on the echo data by using an FIR filter, and filtering random noise in a signal; carrying out ultrasonic imaging based on a convolutional neural network algorithm, carrying out preprocessing according to the obtained scattering echo signal, then taking the preprocessed signal as the input of the convolutional neural network, carrying out imaging on the detected workpiece, and obtaining a coarse scanning image of the detected workpiece; and performing defect edge detection based on a Sobel operator, and performing edge extraction on a bright spot in a final imaging result by using a Canny operator so as to obtain position information