自主式水下机器人（AUV）具有自主导航、自主避碰和自主作业能力，自主作业是水下机器人实现智能化的关键技犬。本文着重叙述AUV对特定作业一“水下目标探测与识别”作业控制和实现过程，叙述水下三维目标的识别机理和识别算法以及为正确识别目标提供最佳条件的预处理[1]、硬件系统和实时处理功能。最后给出试验结果，说明作业目标控制过程和识别算法具有实时性、高效率和可靠性。 Autonomous underwater vehicles (AUV) with autonomous navigation, self-collision avoidance and self-work ability, autonomous operation is the key technology to achieve intelligent underwater robot. This paper focuses on the AUV’s control and implementation of a “underwater target detection and recognition” operation for specific operations, describes the identification mechanism and recognition algorithm of underwater three-dimensional targets and pretreatment [1] to provide the best conditions for the correct identification of targets, hardware System and real-time processing functions. At last, the test results are given, which shows that the operation goal control process and identification algorithm are real-time, efficient and reliable.