The Restoration Of Motion-blurred Images And Application In Electronic Manufacturing Equipment

With the development of semiconductor industry, visual position becomes the core technologies which satisfy the requirements of high precision and speed for electronic manufacturing equipment. High speed movement of vision system contributes to the motion degradation, while high precision requests high-precision image. Therefore, motion blur recovery becomes a key issue for electronic manufacturing industry.Begin with the basic theory of image restoration, this paper briefly introduces image degradation model, classic recovery method and evaluation standard. Motion blur degradation model which provides theoretical basis for subsequent analysis is emphatically introduces. This paper transfers the problem of motion blur restoration into the problem of parameter identification for motion blur. From the spatial domain of motion blurred image, this paper deeply analyses the impacts to the motion-blur image spatial domain and discuss the characteristics introduced by the motion. After the analysis of the image in the electronic manufacturing system, this paper discusses the character of one pixel which has been blurred by motion, and introduces this method to the whole image. Due to the work mentioned above, this paper puts forward motion blur parameter identification method which bases on the second order differential. Considering the noise sensitivity of second order differential, this paper uses a variety of methods to restrain this sensitivity. In order to accurately reflect the method features, this paper respectively conducts comparison test with the methods of this paper and literature in conditions of noise and no noise. Results show that this method has higher robustness and accuracy.This paper identifies the motion blur parameters of motion blurred images in electronic manufacturing equipment with the methods in this paper and literature. Based on these parameters, this paper restores the motion blurred images with wiener filtering. The results show that this method has the great recognition effects for image restoration, satisfies the requirements for visual inspection and positioning, and improves the stability and precision of the equipments.