Research And Implementation Of Character Expression And Background Synthesis System For Film Production

An important way of attracting audience is the posters of film and television works.The exquisiteness of the poster and the unconstrained imagination tend to attract more viewers to watch.However,when the posters are made by the staff in the later stage,a problem often arises: it is difficult to capture the best moment of expressions.Therefore,people will shoot posters several times in order to achieve better results.This kind of work is more than a lot.Many types of special effects need to be added to the background of the poster.These special effects need to be processed in the later stage with certain image processing techniques.This is more technically demanding.On this basis,special effects are laborious and time-consuming.Short videos have taken the throne of Internet dominance.The beautiful background and excellent special effects have attracted more audiences for the short video industry.Therefore,for the later stage of short video,special effects and image processing are also very important.This paper studies the status quo of facial expression synthesis at home and abroad.And the research significance of facial expression synthesis and background synthesis at home and abroad,and analyze the existing problems.Introduced the research on the technology of human facial expressions and poster background synthesis with human or elements,and optimized the application of face synthesis and background synthesis in the production of film and television posters and short videos.The main modules implemented in this paper are the facial expression synthesis module,the facial brightness adjustment module,the background synthesis module,and the video synthesis module.This system can meet the synthesis requirements for most facial expressions,and ensure that the synthesis of the background is natural.This article has studied in depth the problems existing in the synthesis of traditional human facial expressions and special effects background synthesis,and made improvements based on the existing problems.In terms of the facial expression synthesis,using the principle that the relative positions of corresponding feature points are similar when the source face is consistent with the target facial expression.The system is used to calculate the new feature points of the closed organ.This technology is more accurate than the feature point positioning of traditional expression synthesis.Based on the calculation of the new feature points of expression synthesis,triangulate new feature points.In order to obtain the transformed facial expression,an affine transformation is performed on the local triangle.This system does not use deep learning,but uses traditional expression mapping algorithms.The system can complete the synthesis of facial expressions of most people.And using this system can improve the efficiency of function realization.In the background synthesis,this system uses Poisson fusion to seamlessly merge the elements with the original background,so as to merge the existing background with the elements user need.Compared with the matting algorithm,this method has no obvious boundary problem and is more realistic.In the module that combines the characters and the background,the mask is processed by the erosion and expansion operations to make the output result more realistic and fuller.At the same time,the system provides users with options for fusion of different color spaces.Finally,this system combines facial expression synthesis and background synthesis to realize the output work of composing pictures into short videos.