Detecting Significant Changes in Image Sequences

Abstract

In this chapter the authors propose an overview on contemporary artificial intelligence techniques designed for change detection in image and video sequences. A variety of image features have been analyzed for content presentation at a low level. In attempt towards high-level interpretation by a machine, a novel approach to image comparison has been proposed and described in detail. It utilizes techniques of salient point detection, video scene identification, spatial image segmentation, feature extraction and analysis. Metrics implemented for image partition matching enhance performance and quality of the results, which has been proved by several estimations. The review on estimation measures is also given along with references to publicly available test datasets. Conclusion is provided in relation to trends of future development in image and video processing.

Multidimensional Sequence Clustering With Adaptive Iterative Dynamic Time Warping

“…It grounds on the use of both modified iterative dynamic time warping and Kohonen self-organizing clustering map. In contrast to the early works [12,13], the use of both proposed approaches allowed segmentation process in matrix form (more logical in terms of multidimensional data processing) with Frobenius norm, which made it possible to avoid the vectorization/devectorization process to increase the speed, which is especially important in video processing. It should also be noted that the experiments were conducted on more than 20 different videos from 'Destroyed in Seconds' documentary cycle and as a result of clustering, shots with content-based similar characteristics were obtained.…”

“…Aside from video, the problem of temporal sequence segmentation and clustering exists in many practical applications [5][6][7], and a great number of algorithms have been developed by now to cope with this problem [8][9][10][11][12][13]. When the sequences under research are of different length, the situation becomes even more complicated and influences inability of applying traditional metrics for cluster analysis.…”