基于照度调整的矿井非均匀照度视频图像增强算法

发布时间：2018-01-21 06:48

  本文关键词： 非均匀照度 多尺度引导滤波 亮度调整 细节增强参数 照度增强系数　出处：《煤炭学报》2017年08期 　论文类型：期刊论文

【摘要】：由于煤矿井下空间环境受粉尘等因素的影响,视频监控系统获取的作业环境图像存在照度低、照度不均匀等问题,而且现有的算法在处理煤矿非均匀照度图像时会出现颜色失真或者过增强现象,这不利于对图像的判读和应用。结合煤矿的实际数据,提出了一种新的图像增强算法,克服了现有算法存在的问题:分析井下视频图像特点,采用具有边缘保持特性的多尺度引导滤波获取照度分量;基于Retinex理论,将图像分解为照度分量与反射分量;针对照度不均匀的特性,提出一种新的"S型"曲线函数对其进行调整;分析图像的特性,引入受限对比度自适应直方图对其对比度进行增强;提出新的图像增强模型,利用细节增强系数和照度增强系数实现对图像的综合增强。与其他4种算法的对比试验表明,本文算法在主观评价和客观评价方面都优于上述算法。实验证明,本算法具有有效提升图像整体亮度和对比度,同时避免光源附近亮区域的过增强现象的特点,能够满足矿山实际应用需求。
[Abstract]:Because the underground space environment of coal mine is affected by dust and other factors, the image of working environment obtained by video surveillance system has some problems such as low illumination, uneven illumination and so on. Moreover, the existing algorithms will appear color distortion or over-enhancement phenomenon when processing the coal mine non-uniform illumination image, which is not conducive to the interpretation and application of the image, combined with the actual data of the coal mine. In this paper, a new image enhancement algorithm is proposed, which overcomes the existing problems: analyzing the characteristics of downhole video images, using multi-scale guided filter with edge holding characteristics to obtain illuminance components; Based on Retinex theory, the image is decomposed into illuminance component and reflection component. A new "S-shaped" curve function is proposed to adjust the inhomogeneity of illumination. The characteristics of the image are analyzed and the contrast is enhanced by using the restricted contrast adaptive histogram. A new image enhancement model is proposed, which uses detail enhancement coefficient and illuminance enhancement coefficient to achieve image enhancement. This algorithm is superior to the above algorithm in subjective evaluation and objective evaluation. Experiments show that the algorithm can effectively improve the overall brightness and contrast of the image, while avoiding the phenomenon of over-enhancement near the light source. It can meet the needs of mine application.
【作者单位】： 北京大学地球与空间科学学院;
【基金】：国家重点研发计划资助项目(2016YFC0801800)
【分类号】：TD76;TP391.41
【正文快照】： 煤矿视频监控是煤矿安全生产监控系统的重要组成部分,为保障煤矿企业的生产提供了强有力的支持与保障。随“工业4.0”在煤矿行业的推进,“高科技煤矿”信息化建设[m]逐渐成为煤矿从业人员的共识,视频图像处理作为“智慧煤矿”数据采集和分析的基础,将扮演着越来越重要的作用。，

本文编号：1450770