Development of an Improved Feature Based Algorithm for Image Matching

摘要

Image matching plays an important role in many fields such as pattern searching & recognizing[6], image analysis, robotics & computer vision. It is a method to find a certain image in the image database which matches or can be said similar to the given template picture. The template image can be thought of as a subset of the matching image. This paper aims at the improved matching algorithm which is based on the image feature point By searching correct feature point and setting bidirectional threshold value, the matching process can be quickly & precisely implemented with optimistic results. Visual C++ to be used for design and implementation. In future, the feature based algorithm can be modified to choose feature selection threshold adaptively depending on the image's content.