Target detection based on a model of visual attention for UAV

摘要

Visual attention model is a kind of model with good robustness of bionic vision. For ground target detection on UAV (Unmanned Aerial Vehicle) platform, in this paper a target detection method based on visual attention model is proposed, and applied to the complex terrain background for target detection. Using amplitude modulated Fourier transform phase of Fourier transform generate a scale adaptive Gaussian filter. In order to quickly extract the ground targets in aerial images, extracted gradient feature is to detect visual saliency area, and then by using watershed transform method, target image segmentation is achieved. Experimental results show that the method in this paper is capable to be adapting to complex ground target detection, the designed adaptive Gaussian filter is not only de-noise images effectively but also can help reserve original information as possible. In addition, calculation of the proposed method is pretty simple, and so suitable for engineering application.