In order to design a moving target fast tracking system with respect to a timelimited and stable tracking process, especially when the shape of moving objective or its environment condition change, a new approach based on matching local feature points named KPM (key points matching) is first proposed, and local Multiscale feature extraction and matching technology for images are also been researched. First, based on SURF (speeded up robust features) algorithm, interest points and vectors are presented. Second, combine nearest purifying and consistency purifying to move out features outside the target area, so that we can decrease the failed matching and improve the tracking precision. Finally, generate target affine transform matrix and update the moving parameter of the target. Experimental results indicate that KPM is mostly able to achieve a stable tracking while with the monitored target rotating, scale changing, and also the environment illumination glittering. Moreover, it can satisfy the system requirements of tracking stability, higher precision and antijamming.