Research On Key Technologies Of Precision Measurement Based On Image Processing

Precision measurement is a kind of non-contact measurement technology based on image processing technology,which is relatively mature and widely used in recent years. The principle is measuring by detecting the signature of the object's image, which is used as a information carrier for detection and transmission, such as edges.As the rapid development of the production ,science and techology, the measurement accuracy and measurement efficiency have become increasingly demanding, the study of the image measurement accuracy and efficiency becomes increasingly important.Besides,image acquisition process will inevitably produce image noise, which directly affects the subsequent edge detection and image measurement accuracy.This article describes the principle of image measurement and the hardware and software composition of image measurement system,analyzes the source of image noise and its classification, proposes the method of treating the noise points after recognition by comparing the advantages and disadvantages of the traditional denoising algorithm,and then prove that the new algorithm can filter the salt and pepper noises very well under the premise of image sharpness through the MATLAB simulation.This paper analyzes the trational edge-detection operator, and then proves that edge detection based on wavelet transform can get high precision and anti-noise ability through experiments.This paper gets a new sub-pixel edge detection algorithm with wavelet transform for straight line and circle by proposing a new standard of straight line and circle fitting, and then proves that the new algorithm can obtain higher accuracy than the least-squares method under the premise of time-sensitive by MATLAB simulation.