Research On Automatic Editing Method Of Product Presentation Videos

In e-commerce,product presentation videos which vividly display product information have attracted the attention of a large number of merchants due to short length and clear cut functions.Product presentation videos have been the most important method for content marketing after text and pictures.However,the rise of e-commerce not only promotes the development of product presentation videos but also brings a typical problem for video making--the traditional production methods are growing more difficult to cater to the rapidly increasing demand of videos and the fast iteration of product life circle.Editing is an important process for video production,and it has a significant to the quality of the video.The current editing work is done by the editors,which requires time-consuming and labor-intensive processes such as review,assembling,early cutting and final cutting.Therefore,the automatic editing of product presentation videos will increase the efficiency of video makingThe existing work pattern lacks a systematic design framework for automatic editing methods for product presentation video.At present,the video storyboards are based on complete shots or dialogues,which does not meet the demand of product presentation videos that it needs to be compact and display the main features of the products.Shots arrangement is mainly conducted according to the timeline to tell a story or demonstrate actual events,which cannot process the materials of product presentation videos that have no obvious chronic relationship.Most of the algorithms for solving the optimal sequence of shots are performed by limiting the order of shots,which lacks overall constraints.To give solutions to the above issues,the article put forward the opinion that the editing of product presentation video should serve the purpose of enriching information it carries and improving the consistency,based on the theory in consumer psychology that product presentation video can increase the imagination of consumers and reduce their risk perception.It aims to provide a design framework for the automatic edition for product presentation video.Taking garment as an example,the article proposed a new sub-shot segmentation algorithm based on information classification.The algorithm establishes an information classification model which determines segmentation points by comparing the information category of each frame and sliding window.When assembling the shots,the editing rules are coded and a Hidden Markov Model is established,in order to ensure the consistency of the shots.The article also proposed an artificial timeline assemble solution based on the information arrangement logic,established an overall diversity constraint to ensure the content amount of the video,and attempted to have the optimal shot sequence with an improved Viterbi algorithm.Last but not least,the article also showcased the automatic editing system for product presentation videos to help the non-professionals to edit.After comparing the videos edited by professional editors and the automatically edited videos,it is concluded that this method can create consistent product presentation videos with rich information,which verifies the correctness of the theory of this article and the availability of the system.By exploring the automatic editing method of product presentation videos,the article found a new way to increase the efficiency of video making.The method provides a solution to the great demand and rapid iteration of product presentation videos,which is of great significance to the development of the e-commerce industry.