竖井井壁影像的无缝拼接技术研究

摘要

煤矿井筒是煤矿安全生产的咽喉部位,对确保煤矿安全生产有着重要的意义,其安全状况及病害变换趋势可由竖井井壁影像获取.本文针对井壁影像的纹理结构极大相似、影像对比度较差、亮度不足等特点,提出一种基于最小二乘法拟合速度模型的影像拼接方法.首先对影像进行裁剪、亮度增强等预处理,再根据设备及相机的运行参数拟合影像拼接模型,由模型参数建立影像拼接索引,最后分区进行拼接.实验结果表明:该方法既满足了实际需求,又提高了影像拼接在矿井巡检中的实际应用价值,取得了满意的效果.%Coal mine shaft is the throat area of coal mine safety production,and it is of great significance to ensure coal mine safety production.The safety condition and trouble transformation trends can be got from shaft sidewall image.As for the features of sidewall image such as greatly similar texture structure,poor image contrast and lack of brightness,an image stitching method based on the least squares fitting velocity model.First do some pretreatments such as image cropping and brightness enhancement then match image splicing model according to the equipment and the operation parameters of the camera.The model parameters to establish image splicing index,the last partition for Mosaic,next Build image splicing index based on the Model parameters,finally put them together according to the parts.Satisfactory results have been achieved as experimental results show that the method not only can meet the actual demand,but also improve the practical application value of image stitching in mine inspection.