A continuous zoom target recognition system and method with adjustable field of view

Abstract

The invention discloses a continuous zoom target recognition system and method with an adjustable field of view, belonging to the technical field of target recognition. The invention discloses a continuously zooming target recognition system with an adjustable field of view, which comprises a continuous zooming primary subsystem with an adjustable field of view, an image acquisition primary subsystem and a target recognition primary subsystem. The continuous zoom primary subsystem with adjustable field of view is used for precise focus capture of moving targets within the variable field of view. The first-level image acquisition subsystem is used to collect the infrared and visible light of the focused target and generate the target image database. The first-level subsystem of target recognition uses the target image database to train the deep neural network to realize accurate recognition of the collected target images. The invention also discloses a continuous zoom target recognition method with an adjustable field of view. Under the condition that the target moves, the present invention can realize the clear acquisition of the moving target and accurately identify the acquired target.

Description

technical field [0001] The invention belongs to the technical field of target recognition, and in particular relates to a continuous zoom target recognition system and method with an adjustable field of view. Background technique [0002] As a kind of information dissemination carrier, image has the characteristics of large amount of information and intuition, and is easy to be understood and accepted by people. It is an important way for people to communicate with others and the outside world. However, target recognition and information acquisition based on people's cognition of the characteristics of external objects will undoubtedly increase the workload of people. At the same time, the establishment of target feature databases often needs to integrate multiple environments and consume a lot of time, which is not conducive to relying on manpower to establish many targets. Feature database, in order to free people from the monotonous task of target image recognition and fe...

AI Technical Summary

Problems solved by technology

The target recognition system needs to collect infrared and visible light on the target before the target recognition. In view of the dynamic change of the target and the moving field of view, a continuous zoom optical system with an adjustable field of view is required to collect the target first. The traditional infrared imaging zoom process cannot be realized. The field of view is adjustable. For example, the CN204515227U patent "an infrared continuous zoom lens" of the Eleventh Research Institute of China Electronics Technology Group Corporation proposes a refraction continuous zoom lens. adjustment and large aperture; the CN103345059B patent "reflective deformable mirror zoom method and system" of Xi'an Institute of Optics and Fine Mechanics, Chinese Academy of Sciences proposes an off-axis reflective zoom optical system, but the system will cause Due to the problem of image plane movement, it is difficult to meet the needs of adjustable field of view and stable image plane in the target detection process; and the problem of central occlusion will occur when simply combining refraction and reflection to form a coaxial continuous zoom optical system. Therefore, a visual field based on deformable mirror is proposed Field-adjustable off-axis catadioptric continuous zoom system for infrared and visible imaging of moving targets with varying fields of view

Images