Skip to Main Content
This paper presents an unconventional approach for motion detection using 3D image histogram sequence analysis. The idea of this approach is to remove the process of motion estimation into the field of 3D image histogram sequences by analysis of their statistical characteristics. A functional and statistical model of a system for motion detection has been created. This model includes procedures for image processing like: mode filtering for homogenizing the area of objects and its background, defining 3D image histogram sequences as well as procedures for defining statistical characteristics of the 3D image histograms sequences, choice and analysis of criterions for motion estimation. The experimental results prove that the relationship between statistical characteristics of the 3D image histograms sequences and process of motion estimation is a guarantee for creating a reliable and high precision motion detection system as well as for image compression using 3D entropy functions analysis. Such a system can be used for security control of banks, airports, military objects, embassies, shops etc.
Date of Conference: 5-7 Sept. 2005