A Design And Implementation Of The Speed Of The Scenic Image Stitching System

Image mosaicing technique belongs to the category of digital image processing techniques, the research is based on image rendering it is an important task to construct the main technology of virtual environment. Through the image mosaicing technique will be a part of the overlapping pictures or video images seamless and get over wide Angle images or 360-degree panorama image. Panorama in scientific research and engineering projects of strong practicality, widely used in aerospace photography, medical field image processing, virtual scene shows, etc.The thesis analysis abroad panorama generation algorithm, expounds the panorama of the key technology and research status. Through the analysis of the existing algorithms for the camera, the real image, the thesis realized the stitching characteristics of a high-speed panorama stitching algorithm. Mainly realizes the panorama panoramic and two functions, with multiple around the panorama view chart analysis for the research target. Users can be two pieces of overlap and image mosaicing travel.The proposed algorithm is thinking: In this thesis, a fast matching method based on traditional methods, designed a landscape of high-speed image mosaic algorithm.For the overlapping part of the outdoor scenery images to overlap the gray block, the average density of the standard deviation to measure whether the registration of this regional mosaic.By calculating the horizontal and vertical stitching distance to overlap two images pixel weighted average position of image overlap, and then fused.Complexity of the algorithm is o (image height + image width), experimental system based on computer vision library opencv 2.0 image processing, after experimental verification, the algorithm has better results.