Electronic image stabilization based on improved gray projection

Gray projection method for electronic image stabilization has poor image stabilization effect and it may cause the error stability when there is a moving target in the image sequence. In order to solve the problem, we proposed an improved gray projection algorithm which referred to the position estimation method for partial moving target and the pixel alternative method, based on the analysis of the basic theory for gray projection. We estimated the position of the partial moving target using a change of gray value, and removed the partial moving target with the method of pixel alternative to stabilize image. Through analysis and verification, the conclusion is that this method can well process the image which has problems and improve the image stabilization effect. Using the improved method, the peak signal-to-noise ratio (PSNR) is increased by 1.35.