| Image mosaic means making seamless mosaic for the images with connectivity, and producing an image with high resolution and wider visual angle. Usually the mosaic images have overlapping parts, and the mosaic image should be as similar as original images, and the distortion is as little as possible without clear splicing lines. As an increasingly popular technology, Image Mosaic has wide application in various areas such as Virtual Reality Technology (VRT), computer graphics, medical image processing and so on.In this paper, we discussed main theory and technology of image mosaic. Several methods of image mosaic were discussed. Firstly, we studied the template registration algorithm, which is based on the pixel information of the image. Then we turned to another method which bases on image characteristics. We took the Harris corner algorithm as an example to carry on a Babu photo processing simulation. After that, we studied a third algorithm, which is based on the frequency domain information. We discussed its feasibility and performance.Image acquisition in the Babu producing line will be affected by the specific environment of the scene. The images, got by the CCD and video capture card, had problems such as noise, geometric distortion, intensity changes and so on. For the adverse effects coursed by external factors in image registration, this paper introduces a check-style template matching algorithm. Then we use template information, rather than itself, to calculate the correlation of the two templates. The false matching were prevented with low costs. This kind of algorithm has high speed and simple form, and is easy for batch processing and online real-time processing. It is suitable for real-time image processing on the Production line.This paper uses matlab simulation to discuss advantages and disadvantages of different algorithms. The experimental results show that our algorithm is better than the original algorithm in accuracy and speed. |
PDF Full Text Request