Computing Images Invariant Characteristics with Chain Coding
Zaw Win Htet, Victor D. Koldaev · 2018 International Russian Automation Conference (RusAutoCon) · 2018
The paper investigates the properties of invariant moments of binary images, which is necessary for their formation and analysis in the problem of pattern recognition of graphic images. It is shown that the invariants have different sensitivity to the changes in the input data, which determines the strategy of their choice. Invariant Xy moments are used for the pattern recognition and distance measurements. The paper discusses the method of calculation of invariant characteristics for the binary images. The results of experimental determination of the quality of recognition are presented by their contour images of the various objects, as well as the results of the comparison of recognition using different sets of invariant moments. The proposed approaches to object recognition based on the method of chain coding can be provided the reduction of the amount of data and can be used for image recognition in the technical vision systems.