I.V. Zhigulina

Algorithms of recognition of moving objects are necessary at creation of automatic image processing systems. In this article the problem related to the automatic recognition of moving objects is considered. This problem is divided into three stages: automatic finding of a movement, the correction of a smearing, and the recognition of objects. Phase-energy partial spectrum and partial energy spectrum are used for automatic finding of object-s movement. Division of the image into parts increases probability of correct detection. The one-dimensional model of a smearing is gained and the correction's algorithm is developed. The bank of S-matrixes may be used for the algorithm of recognition of object's moving, instead of the gang of standards. This algorithm is applicable at the third development cycle of automatic system of comprehensible complexity working in real time.