Methods to reduce network load when transmitting computer images in real time

Abstract

The new approach related to the reduction of network traffic when transmitting computer images in real time has been studied. Two methods for computer image encoding and transmission in networks have been proposed. First method is directed to full-color computer image compression and transmission based on color palette reduction and "color coarsening", dictionary encoding and the second one aimed at computer binary compressed image transmission with the use of adaptive statistical threshold, based on the iterative classification and segmentation using line primitives and object boundaries to obtain binary image representation. The comparative results of four experiments of full-color image representation; grayscale image; image changes and color indexation; image changes with binarization (last three with data compression) have shown the effectiveness of the proposed methods for data transmission with several hundreds of times compression ratio depending on the image type and content in comparison with original image.