Bionic navigation algorithm of contrast entropy flow

• 1、National Key Laboratory of Science&Technology on multi_spectral information processing,Huazhong University of Science and Technology,Wuhan,430074

Abstract：Existing navigation methods have some shortcomings, flying insects can achieve accurate navigation in complex natural environment, therefore doing research about the mechanism of flying insects can help to develop a more efficient navigation system. Visual navigation experiments of flying insects (bees represented) show that their visual navigation is based on image information flow. Entropy is a representation of information, so the image entropy can be used to characterize the image information. This paper introduces the concept of entropy image and the construction method of it. In order to better characterize the image texture features and spatial distribution, the concept and construction method of contrast entropy image was proposed. This paper refers to the idea of optical flow algorithm and its application to motion estimation to calculate the contrast entropy image and to estimate the motion parameters, then combines with Kalman filter to optimize the results. Experiments show that the contrast entropy image navigation algorithm has better performance than the gray entropy image navigation algorithm.

Keywords： Navigation guidance and control bionic navigation entropy image contrast entropy image entropy flow Kalman filter