CRITERION AND ALGORITHM FOR DETECTING DYNAMIC OBJECTS IN A COMPLEX BACKGROUND BY A LOW-CONTRAST POINT IMAGE
A. N. Katulev1, A. A. Khramichev1, O. V. Guzenko2
1Scientific Research Center of the Central Scientific Research Institite of Aerospace Defense Forces of the Ministry of Defense of the Russian Federation, Naberezhnaya Afanasiya Nikitina 32, Tver, 170026 Russia
2Military Scientific Committee of the Armed Forces of the Russian Federation, pr. Vernadskogo 100, Moscow, 111399 Russia
Keywords: point image, low contrast, complex background, similar conditions, DO detection criterion
Subsection: ANALYSIS AND SYNTHESIS OF SIGNALS AND IMAGES
Abstract
A criterion and an algorithm of detecting dynamic objects (DOs) in a complex background formed by an intense cumulus and high-altitude cumulus are proposed. The object image has a small size (point image) and low contrast. The principle of DO detection is fractal-correlation: it is based on the use of sampling as a relationship of likelihood functions of similar alternative conditions: either “only complex background within the sight of an optoelectron device (OED)” or “DO on the complex background within the sight of an OED.” The DO detection algorithm is designed as a binary accumulator according to the most powerful local criterion. The critical limit of decision making is defined by the Neumann - Pearson lemma for the acceptable possibility of false detection of a DO. Simulation proves the algorithm to be highly effective.
|
