Machine Vision-based Distance Study Of Tracking Target

Machine vision has rich information. And it's with high intelligent level. Because of those characteristics, the vision tracking and ranging has been applied extensively in a lot of fields, such as aerospace, aviation, navigation and ground. The navigation system based on this technology has some advantages, including small volume, low cost, high autonomy. It has been successfully used in low altitude aircraft navigation, Unmanned Aerial navigation and probe landing navigation. Nowadays, the theory is gradually getting ripe.This paper taking the distance tracking based on machine vision as the object of study, comprehensively used dynamic image processing, real-time target tracking and detecting, and monocular vision-based dynamic ranging technology, completely acquiring the distance between motion objects and the camera, which provides accurate and reliable data to target real-time surveillance.There are five parts in this paper. This main content of this paper is as follows:1. This paper introduces functions of the vision ranging system based on vision tracking briefly, then summarizes the design flow of this system and hardware scheme, as well as the main structure characteristics of DSP and its advantages in digital signal processing. At last, compiles the system flowchart and general software design scheme combining C language programming.2. In this paper, I adopted my improved median filter with smaller calculation amount to preprocessing the images by analyzing the existing preprocessing methods and the tracking system's characteristics.3. Then I summarized the available camera calibration methods, as well as combining the system characteristics, adopted template matching to tracking motion objects and used template updating strategy to update the information of image color and the objects'edges in order to make it adaptive to large image deformation.4. Gray threshold segmentation and edge detection are commonly used image segmentation methods. Gradient threshold edge detection could detect edges completely and suppress noise effectively.5. Ranging used geometric distance measurement method based on camera calibration with smaller errors in the field of view (FOV).In the windows XP system, successfully simulated and tested the algorithm using VC++ programming.