As a mineral resources powers, downhole information monitoring has been country attaches great importance to. Screen monitor is an important way of monitoring. But by the conditions of the underground environment, it is difficult to directly obtain wide field and high resolution images. Image stitching technology can effectively solve this problem. It is easy to obtain high resolution downhole panoramic images.More factors exist in underground environment, fore example weak light and dust. Analyses the classical algorithm, SURF(Speeded-Up Robust Features) is selected as the feature point extraction algorithm. It has the good robustness. In order to reduce the computation, improve the accuracy of feature point matching, cosine similarity and clarity is used to initial screening. Its purpose is to remove the smaller feature similarity. Then the Euclidean distance is used to match points. Finally, the bilinear interpolation and the weighted average method is used to process images. The final image is obtained without seam.The experimental results show that this method can effectively stitch mine image. Final mosaic image is clear and no obvious traces of stitching. |