Computer image retracking is a classical difficult problem in computer image processing. The development of computer image retracking techniques not only influences on other techniques in computer image processing, but also plays an important role in the analysis of biology computer image. Traditional computer image retracking techniques have many defects for complex computer image retracking. Most exiting computer image quality assessment approaches evaluate the difference between computer

more »

... es of the same size and are not suitable for computer image retracking. In this paper, we propose a novel Content Aware Assessment Metric based on pixel tracking to access the quality of computer image retracking. This metric evaluates the similarity between a retargeted computer image and the original regardless of the difference in computer image size or aspect ratio. By tracking the position of every pixel from the original computer image to the retargeted, including both cumulative distance change and cumulative angle change to measure the local difference between the two computer images. By comparing our experimental results with user surveys, we demonstrate that the proposed assessment metric is consistent with human experience.