Objectives  In order to meet the requirements of underwater search and rescue, exploration, target detection and tracking,

  Methods  we carry out a target identification test for the developed "T-SEA Ⅰ" AUV, through which a typical frame is selected for the target detection and identification from the acquired moving target images by virtue of the front-view sonar; describe the system composition and working principles of "T-SEA Ⅰ" AUV; then propose a framework for a control system in which laptop and surface control box act as a surface control unit, PC104 industrial control board acts as the main control unit of an autopilot and the PC104 camera board processes the camera images separately; use three-frame difference method to detected the sonar target, and extract the target characteristics according to a fast ellipse detection algorithm based on Hough transform.

  Results  The lake test shows that the AUV runs reliably and normally and meets the requirements of underwater work; the sonar target identification is good.

  Conclusions  This AUV's control system structure and target identification method can provide a reference for the development of medium and large-scale AUVs and the target identifications in other fields such as video, infrared, etc.