VLSI Architectures for Video Salient Object Detection using 3D Dual Tree Complex Wavelet Transform

Detection of Salient objects in image and video sequences are very interesting and rewarding task which is the crux of many applications that deal with computer vision. Salient objects are segmented from the objects of interest and are correlated with objects in its neighborhood which is similar to the cognitive property of human visual system. Major distinction of using complex wavelets in the process of salient object detection lies in the fact that the complex wavelets have the directional selectivity property. Systolic Array and Distributive architectures are considered for the detection of salient objects in the video sequence using 3D DTCWT. The proposed work are implemented on SPARTAN 6 FPGA and tested in the XILINX.