煤矿鱼眼畸变图像校正

发布时间：2018-07-11 09:47

  本文选题：煤矿监控图像 + 鱼眼图像　； 参考：《工矿自动化》2016年07期

【摘要】：针对煤矿井下视频监控系统中鱼眼图像的畸变问题,根据理想球面透视投影模型的保线性约束条件,提出一种面向内容的校正参数自动获取方法,即通过对鱼眼畸变图像进行数学形态学和随机霍夫变换预处理得到所需的图像边缘特征直线,然后统计水平和垂直线段在总线段中的比例,分析统计结果,获取最佳校正参数。为了减少图像信息过多对校正参数获取过程的影响,对理想球面透视投影模型进行了改进,提出一种改进的鱼眼畸变图像校正方法,实现了对图像中心的转移和部分区域的放大。实验验证了该方法对鱼眼畸变图像的校正效果较好。
[Abstract]:According to the linear constraint condition of ideal spherical perspective projection model, a content-oriented automatic method for obtaining correction parameters is proposed to solve the distortion problem of fish-eye image in underground video surveillance system of coal mine, according to the linear constraint condition of ideal spherical perspective projection model. The image edge feature line is obtained by mathematical morphology and stochastic Hough transform preprocessing of fish-eye distortion image. Then the proportion of horizontal and vertical line segments in the bus segment is calculated. The statistical results are analyzed and the best correction parameters are obtained. In order to reduce the influence of too much image information on the process of obtaining correction parameters, the ideal spherical perspective projection model is improved, and an improved correction method for fish-eye distortion image is proposed. The transfer of image center and the amplification of some regions are realized. Experimental results show that the proposed method is effective in correcting fish-eye distortion images.
【作者单位】： 华晋焦煤有限责任公司;中国矿业大学信息与电气工程学院;
【基金】：国家自然科学基金资助项目(U1261105) 徐州市科技计划项目(KC14SM099)
【分类号】：TD76;TP391.41
，

本文编号：2114692