Error Analysis And Accuracy Improvement Of The Strain Measurement By Video Image

As a modern measuring technique, image measurement technology is the basis of modern optics, financial optoelectronics, computer graphics, information processing, computer-vision technology. Image measurement is non-contact, high-speed, dynamic range, informative, relatively low cost, and has many other advantages. Compared with traditional measurement methods, image technology has obvious advantages. But a key issue that limited the use of image measurement technology is the accuracy, reliability of the image measuring, just the measurement accuracy.In order to increase the accuracy of the image measurement system, and to accelerate the popularization and application of image technology development, we researching on the subject of the image in the light of The video strain measurement system (video-extensometer) in the process. Deeply we researched on how to improve the accuracy of computer graphics. In particular, we examined the software algorithms to improve accuracy; researched The video precision strain measurement system used to the testing machine; designed the components of Video-extensometer system, and hardware and software environment for algorithm and the match with the test machine to achieve the improvement of measurement accuracy. The main contents of this text as follows:1. First, the text introduced the composition of the image strain measurement system. Then the main issues affecting image accuracy from hardware factors, such as lighting field noise, thermal noise electronic, CCD performance, lens distortion, quantization error, temperature, vibration, the effects of video feeder. The text analyzed the images strain measurement error of the measurement accuracy, gave an effective way to eliminate or reduce errors. Image measurement so as to improve the measurement accuracy is specified in the right direction.2. Using software to improve the measurement accuracy is a simple and effective method, and costs equipment low. We research Image Measurement System software algorithms. We collected a detailed study on the theories of evaluating image quality, Sub-pixel edge detection and location, edge fitting and follow-up data filtering, which can improve the accuracy of measurement. We completed the methods of the evaluation of image quality, Sub-pixel edge detection and accurate positioning data filtering, edge fitting and follow-up data filtering. It effectively improved the accuracy of the video strain measurement from software algorithms.3. Image measurement system eventually applied to the actual measurements. Measurement precision is not only by the system's hardware and software limitations, but also limited by the outside world testing environment and the object which is measured. Such as vibration, light, the light of the measurement system, workpiece material impact on the measured temperature. This requires measuring system with the surrounding environment, equipment and other reasonable explanation for matching to ensure the accuracy of measurement system. Based on research topics of the video strain measurement system, the text finished matching the video strain measurement systems with test machine, completed a reasonable mix of data sharing and communication, which ensure the accuracy of the measurement system as a whole.4. At last the text analyzed the experimental of the image measurement system. We verified the methods to improve accuracy, including sub-pixel algorithms, the fitting algorithm comparison, the overall experimental results analysis. Then we selected the most effective ways to improve measurement accuracy. It analyzes the problems and shortcomings, the accuracy of the measurement system could further enhance the image of options. From the experiment, under certain conditions, the accuracy of the image strain measurement system can be achieve require.The paper is just research on improving image precision method for the study of the image strain measurement system, that's relative. Improve the accuracy of image measurement system is a huge and complex project, need to in consideration of various factors. This paper just seeks to use some methods to improve the accuracy of the image strain measurement system, hoped to play a little help to the development and application of image technology in domestic.