Improved Electronic Image Stabilisation Based on Image Mosaic and Grey Projection
Shaoqing Tian · Review of Computer Engineering Studies · 2017
Smart devices are being used more and more in people's daily lives, such as iPhone, Android, UAVs and Google Glass.These devices all have one or more cameras, with sophisticated camera functions and as a result, a large number of images are produced, including continuous and discontinuous, static and dynamic, local and panoramic, blurred and clear images.These images appear due to the functions, performance or operation of the devices or the users' own needs.Although these images keep people's beautiful moments or wonderful processes, not all of them can be easily obtained because some or all of them have problems, whether manmade or non-manmade.Therefore, it is necessary to study the problems existing in the process of image acquisition, including image stabilisation, panorama image acquisition and rapid graphics processing, so that people can better use smart devices to capture their desired images.Based on these problems, this paper presents an electronic image stabilisation method based on image mosaic and grey projection to address image losses after electronic image stabilisation.In addition, a main contribution of this paper is using image mosaic to improve the processing results of the electronic image stabilisation algorithm.Finally, we verify the algorithm based on grey image projection. RELATED LITERATURE REVIEWThere are many related research results although few scholars and experts have contributed in this regard.In terms of the Electronic Image Stabilisation (EIS) technology, the research trend has always maintained stable, indicating that this technology has entered a mature stage, so currently scholars in this field mainly focus on combining multiple algorithms and application.Specific researches are as follows.Zhao [1] described the visual characteristics of human eye, and proposed a kind of inter-frame motion amplitude statistics algorithm based on the video quality evaluation algorithm and human's visual characteristics.Being in high consistency with the subjective evaluation results, this algorithm can effectively evaluate the performance of the electronic image stabilisation algorithm.Fan [2] studied the mine car camera system.In order to quickly and accurately stabilise the jittered images obtained from the system, an electronic image stabilisation algorithm based on downsampling grey projection is proposed.This method is superior to the traditional grey projection algorithm in terms of accuracy and computation time.Gong [3] described the shortcomings of the grey projection algorithm, such as single grey value and poor contrast, and proposed an electronic image stabilisation algorithm based on block grey projection, which can improve the peak signal-to-noise ratio by 21.5%.Bai [4] studied the application of the block grey projection algorithm in satellite assembly and proposed an electronic image algorithm based on partial grey projection to ensure that the robot system could output stable and coherent satellite assembly images.Yuan [5] analyzed the slowness and inaccuracy of the traditional estimation methods under video jitter and proposed a motion estimation method combining grey scale and block matching to improve the computational speed and estimation accuracy of the global motion estimation method.Ji [6] described the problem that the feature points in the foreground moving objects would