Study Of Small Detail-amplification Technology Based On Regional Video Image Sequence

As the increasing development of science technology, video image informationis playing an important role in various fields. Resolution ratio is confined as a resultof the limitation of physiology, psychology and nervous system, so some details,whose texture information can make the video clearer and colorful from somecertain imagine enhancement technology, conceal from invisible area or some fieldswhich eyes can not find. Amplifying the motion information of details reflects thepotential information, revealing the important hidden rules.Various image processing technologies are mostly based on denoising andincreasing contrast visible areas, gaining and compensating distortion of details, butthere is little research on invisible area. The thesis, through analyzing and contrasting,finds a latest adaptive detail‐image enhancement technology, applicable toamplifying small texture. The technology, through multi‐scale image analysis,adopting new image fusion technique and edge processing methods, makes smalldetails which people can not see visible. The result of experiment, adopting theenhancement algorithm, is effective and applicable to video image sequence.Moreover, the thesis analyzes some theories and technologies of moving objectsextraction and amplification in small motion, applying to vehicle tracking experimentextracting background‐attachment. Finally, the thesis deepens the research ofmoving objects amplification technology in invisible area, analyzing the latest smallmotion amplification technologies, which are based on LaGrange description andEuler description. It describes the differences between those descriptions,contrasting the results of the experiments, analyzing the merits and defects of thosetwo descriptions. Final integration advantages of these two methods, to find a newimproved algorithm.To sum up, the thesis researches small detail‐amplification technology of videoimage sequence systematically, studying and experimenting the thought, algorithmand theories of small texture and motion, in order to exploring significantly in thisfield.