Image quality is an important factor that influences the dynamic target information perception; it is the key factor of real-time target state analysis and judgment. In order to solve the multi-observation station comparison and video optimum seeking problem in the process of target information perception and recognition, an image quality assessment method based on visual characteristics is proposed for infrared target tracking. First, it analyses the basic infrared target image characteristics and application requirements, analyses the status and problems of the multi station optimum seeking technology. According to the expected research results, the processing flow of image processing is established. Then, the image quality objective assessment index is established, which reflects the basic characteristics of the target image, and the assessment index is integrated into the normalized assessment function. According to the quality assessment function, the infrared image quality assessment based on infrared target recognition and image analysis processing is realized, which is mainly characterized by the region of interest and dynamic visual characteristics. And on the basis of this technology, the real-time optimum seeking of multi station infrared target tracking image is completed. In order to verify the effectiveness of the method and the practical application effect, it designs the quality assessment and comparison of different station infrared images. Example shows that the method proposed in this paper can realize multi-observation station infrared image assessment comparison, image quality sorting, the optimum seeking of the infrared image based on the quality assessment. The results accord with the characteristics of infrared target image and dynamic visual characteristics.