The Research On Object Matching Based On Major Color Spectrum And Spatial Distribution Entropy

Due to the urgent needs in many fields such as intelligent transportation, medical treatment and security, multi-camera monitoring techniques including object detection and location, object matching, object behavior analysis and so on, have been paid high attention. And object matching as the bridge of multi-camera monitoring has also become the keen and difficult topic in this field.The methods of object matching are the main work in this thesis. Firstly some common object matching methods are introduced and the methods based on the color characteristic are analised further more. And then, some problems of the methods based on color are pointed out, such as a large amount of calculation, insufficient color spatial distribution, inaccurate matching method and so on. The main contributions for the thesis are as follows:(1) A major color spectrum description method based on fusion of the second class center optimizing and M-Kmeans is proposed. The second class center optimizing is used to choose the start points of the Kmeans clustering for obtaining color spectrum to reduce the amount of calculation, and M-Kmeans is used to adjust the members in the clustering process. Research and experimental results show that the fusion algorithm not only improves the accuracy of the description of the color spectrum, but also reduces the sensitivity of the clustering results to the start points, and improves the stability of the clustering results.(2) The technique about spatial distribution entropy of major color spectrum is proposed. According to the proportion of the distributing pixels in the different zoning, the entropy of the major color spectrum can be calculated based on Shannon entropy formula, which represents the spatial distribution information of the object appearance color. The experimental results show that the entropy technology could distinguish the objects with similar major color spectrum but different color spatial distribution.(3) Based on the results of the first two studies, The MCS-SDE object matching method is proposed. The similarity of major color spectrum is weighted by the similarity model of the entropy of color spatial distribution while calculating the object similarity. More ever, multi-frame matching method is discussed at the mean time. The experimental results show that the new object matching method increases the discrimination between the matching pairs of objects and un-matching pairs, at the meantime, it is easier to determine the threshold of judgment than the method in the reference [11], and obtains more accurate and robust results at the meantime.Finally, the summary of this thesis is made, and the future study direction on this field is given as well.