A method and device for automatic target recognition of millimeter wave images

Abstract

The invention discloses a method for automatic target recognition of a millimeter-wave image, comprising the following steps: (1) acquiring a sub-image block of a millimeter-wave image to be identified; (2) acquiring the sub-image based on a trained convolutional neural network The block contains the probability value of the target; (3) based on the probability value, the probability cumulative map of the millimeter-wave image to be recognized is obtained; (4) the target is marked based on the probability cumulative map, so as to complete the millimeter-wave image to be recognized target recognition. Correspondingly, the invention also discloses a millimeter wave image automatic target recognition device. The millimeter-wave image automatic target recognition method and device described in the present invention are suitable for automatic target recognition of millimeter-wave images, have a good target recognition effect, and solve the problem that it is difficult to obtain good results for millimeter-wave images using traditional manual design features in the prior art. The problem of the detection effect is solved, the target is accurately positioned, and false alarms are reduced, the security inspection efficiency is improved, and the labor cost is reduced.

Description

technical field [0001] The invention relates to the field of imaging security inspection, in particular to a millimeter wave image target recognition method and device. Background technique [0002] In the field of security checks, commonly used security checks include metal detectors and X-ray security checks. Metal detectors can only be used to detect metal objects, and cannot detect non-metallic dangerous objects such as ceramic knives; X-rays can be used to detect metal objects, but they are ionizing radiation like nuclear radiation, which is easy to cause harm to the human body and can only be used as luggage security check. Due to the above defects, traditional metal detectors and X-ray security checks are not ideal human security checks. Millimeter waves can penetrate clothing and detect prohibited items such as guns, ceramic knives, and unknown liquids hidden under clothing, and millimeter wave radiation is non-ionizing radiation, which has low power in actual use ...

AI Technical Summary

Problems solved by technology

Due to the instability of the millimeter-wave imaging system, the millimeter-wave human body images often have high noise, and the body shapes of the scanned people are different, and the textures of different prohibited items vary greatly. It is difficult to obtain good detection results by using traditional manual design features.

Images