A yield and reliability enhancement framework for image processing applications

Abstract

How to improve yield and reliability of an electronic system has been a challenging problem, especially for mission-critical applications such as automotive electronics. In the literature many image enhancement methods have been developed to increase the quality of a given image whenever needed. However these methods mainly deal with unsatisfying yet error-free images. In this paper we present a yield and reliability enhancement framework that can cope with both fault-free circuits and faulty ones by combining image enhancement and testing methods. By carefully evaluating image quality, the proposed framework can not only improve the quality of an error-free image, but also reduce the error significance of an unacceptable erroneous image so as to maximize their acceptability in a certain application. The experimental results show that minor vibrations in images are almost insensible and one can easily increase the acceptability of an image using an appropriate image enhancement method.