Abstract: In order to monitor the production of underground in real-time through image acquisition technology, underground image stitching algorithm based on image enhancement is proposed. This paper uses the local bilateral filtering algorithm to enhance image, reuses approximate Hessian matrix and frame-like filtering to determine the position of the feature points, and then calculates the descriptor vector of feature point, using match method of ratio of the closest distance and the next closest distance to match feature points and eliminate error matching by RANSAC algorithm, and finally makes use of the feature points to get the transformation matrix of calculation, uses the method of a linear gradient fusion to realize image fusion. The experiments verify the effect of enhancement algorithm and prove that combining this algorithm with SURF can improve the efficiency of image mosaic by the number of feature points extracted from original image and enhanced image. This is helpful to improve the accuracy of matching and the quickness of splicing.

Key words: image mosaic, image enhancement, Speeded Up Robust Features（SURF）

摘要： 为了通过图像采集技术实时监控煤矿井下生产情况, 提出了基于图像增强的井下图像拼接算法。利用局部双边滤波算法对图像进行增强，在此基础上再利用近似的Hessian矩阵和框状滤波确定特征点的位置，然后，计算特征点的描述子向量，采用最近距离比次近距离的匹配算法将特征点配对，最后利用特征点对计算得出变换矩阵，采用线性渐变融合方法进行图像融合。通过图像增强前后特征点数量对比实验验证了增强算法的有效性，并证明了该算法显著提高了SURF（Speeded Up Robust Features）算法的拼接效率，有利于提高匹配的准确性和拼接的快速性。

关键词: 图像拼接, 图像增强, SURF算法