TY - CHAP A1 - Bohn, Gunther A1 - Möhringer, Peter A1 - Kőrösi, Adam A1 - Mitesser, Oliver A1 - Hovestadt, Thomas T1 - Algorithms for flight path tracking of butterflies at inhomogeneous backgrounds T2 - Papers of 33rd International Scientific Conference "Science in Practice" N2 - In this paper we describe the image processing algorithms for the tracking of the flight path of a single butterfly Polyommatus icarus in its natural surroundings. This tracking allows biologists to analyze the movement decision rules of the butterfly. Our butterfly video-screening and tracking system consists of a matrix of 6 stereo camera systems. The screening area is 4x3 meters. The tracking begins with a preprocessing algorithm, which is robust against the inhomogeneous background and optical distortions caused by the natural environment like moving shadows, change of illumination and color: The preprocessing algorithm works with grayscale images to eliminate the influence of the color changes. It adjusts the average intensity of consecutive images and calculates their difference image. So the influence of the inhomogeneous background and the shadows is greatly reduced. The reprocessing is followed by algorithms for thresholding, temporal tracking and calculation of the 3D-position using two corresponding images of the stereo camera systems. Results show, that the algorithms work robustly in spite of the inhomogeneous background and the optical distortions. The tracking algorithm generates the flight path of one butterfly in the screening area of one selected stereo camera system. A tracking of the butterfly movement is technically feasible over multiple stereo camera systems and so over large areas. KW - imagine analysis KW - path tracking KW - butterfly KW - background subtraction Y1 - 2015 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:863-opus-2586 SP - 50 EP - 54 ER -