The method of rapid evaluation of agrotechnical indices of cotton harvesting machines, based on computer processing of digital images of cotton row before and after machine harvesting, is proposed in the paper. The algorithms and software to solve the problem of assessing the quality of cotton harvester performance and to determine the indices of the agricultural background of the cotton field, based on the analysis of the plot images taken from different angles, are presented. The application of the proposed technique allows significant reduction of time and money to test the machines, rapid evaluation of their performance directly during the tests. The software and hardware developed can be used in test automation, in automatic systems for monitoring and controlling the technological processes and modes of operation of agricultural machinery, in particular, cotton harvesting machines.

No abstract

The problem of pattern recognition in condition of huge dimensions of features' space is considered. Extended model of recognition algorithms on the base of estimates' calculations algorithm is proposed.

No abstract

In this paper, the problem of constructing hybrid algorithms for identifying a person (IAP) by the image is considered, which is the main one in the development of biometric systems based on them. To solve this problem, we propose a model of algorithms, consisting of the following stages: selection of the area of the face in the image; searching pupils on the face image (FI); determining the location of the mouth and nose; forming a set of geometric features; allocation of subsets of strongly coupled features; the definition of representative features in each subset of strongly related features; interim assessment for the FI class for each set of representative features; estimate for the FI class for all intermediate estimates. A distinctive feature of the proposed hybrid algorithm (HA) is the integration of the results of various methods for determining anthropometric points at the level of calculating the final score for each class of facial images. The main purpose of this paper is to develop an algorithm for integrating various methods of identifying features when IAP on a FI.