Research On Panoramic Image Mosaic Technology Of Android Mobile Phone Based On OpenCV

Twenty-first Century is the information age.People are surrounded by various kinds of information every day.90% of these information comes from images.In many situations of life,people need to record some scenes.The commonly used tools are cameras,including smart phones with pictures.With the development of science and technology,software and hardware are constantly updated and upgraded,most of the mobile devices have amazing powerful performance,so the image processing technology that can only be carried out at the PC side in the past because of limited performance,can be turned to the mobile platform,which greatly stimulated the creative enthusiasm of the research workers and the major mobile phone manufacturers.The panoramic mode of mobile phone photography is a typical application of image mosaic technology.Its main goal is to combine some adjacent scattered pictures into a complete and wide field of vision and high resolution images.Today,with the popularity of deep learning and artificial intelligence,panoramic image mosaic technology has wider application market and higher commercial value.After focusing on the basic theory and key technology of panoramic image mosaic technology,aiming at the requirement of real-time application,this paper proposes an improved algorithm for image mosaic based on SURF feature points.First,the SURF feature detector is used to extract the SURF feature points of the input image,and the data structure of the K-D tree is used to save the image.The process of image registration is divided into two steps: initial matching and final match.The initial match uses the BBF algorithm to carry out cyclic search.The result of feature point matching is purified by RANSAC algorithm,and the feature points of mismatch are reduced to the maximum degree.After the final image matching result is obtained,the cylindrical projection is used.By placing the image into the same coordinate system,the weighted average fusion algorithm is used to produce the panoramic image that conforms to the visual consistency.This paper uses the open source computer vision library OpenCV to program and test the algorithm at the PC end.After the debugging and optimization,the algorithm is transplanted to the Android cell phone,and the panoramic image splicing software of the Android platform is developed by using the combination of Android Studio and OpenCV4 Android.Experiments show that the algorithm has good stitching effect in common scenes in life,and has good stability.It can meet the requirements of real time and meet the practical application requirements.