采用图像直方图特征函数的高速相机自动曝光方法
发布时间：2019-02-13 07:14
【摘要】：提出一种采用图像直方图特征(Histogram Feature,HF)函数的自动曝光方法,用于在背景光照快速、大范围变化的情况下对高速相机进行自动曝光控制。首先,采用多点测光对获取的图像进行兴趣区域(Region of Interests,ROIs)提取,以降低系统测光的计算量。然后,通过计算兴趣区域的HF函数选取大步长对曝光时间进行粗调。最后,通过模糊逻辑计算出曝光时间的精调步长并运用阈值限制实现变步长搜索最佳曝光时间,提高高速相机自动曝光的准确性及稳定性。实验结果表明:本文方法可以在2ms内完成一帧图像的亮度测量并对曝光时间进行调整,相较于基于平均亮度值的传统自动曝光方法,在0~110ms内,光源光照强度从760~23 100lux反复变化时,本文方法获得的图像信息熵比平均亮度值法的信息熵均值提高48.38%,方差降低62.13%,可以提供更好与更稳定的图片细节信息,为后续的自动调焦、图像识别以及目标跟踪提供参考。
[Abstract]:An automatic exposure method using image histogram feature (Histogram Feature,HF) function is proposed, which is used to control the automatic exposure of high speed camera under the condition of fast background illumination and wide range variation. Firstly, the region of interest (Region of Interests,ROIs) is extracted by multipoint photometry to reduce the computational complexity of the system. Then, the exposure time is coarse adjusted by calculating the HF function of the region of interest. Finally, the fine step length of exposure time is calculated by fuzzy logic, and the variable step size is used to search for the best exposure time by using threshold limit, which improves the accuracy and stability of automatic exposure of high speed camera. The experimental results show that this method can measure the brightness of a frame of image in 2ms and adjust the exposure time. Compared with the traditional automatic exposure method based on average luminance value, in 0~110ms, the illumination intensity of the light source changes from 760,23 100lux. The entropy of image information obtained by this method is 48.38 higher than that of the average brightness value method, and the variance is reduced by 62.13 points, which can provide better and more stable image detail information for subsequent automatic focusing. Image recognition and target tracking provide reference.
【作者单位】： 中国科学院长春光学精密机械与物理研究所;中国科学院大学;
【基金】：国家863高技术研究发展计划资助项目(No.2011AA8082035)
【分类号】：TP391.41
本文编号：2421308
[Abstract]:An automatic exposure method using image histogram feature (Histogram Feature,HF) function is proposed, which is used to control the automatic exposure of high speed camera under the condition of fast background illumination and wide range variation. Firstly, the region of interest (Region of Interests,ROIs) is extracted by multipoint photometry to reduce the computational complexity of the system. Then, the exposure time is coarse adjusted by calculating the HF function of the region of interest. Finally, the fine step length of exposure time is calculated by fuzzy logic, and the variable step size is used to search for the best exposure time by using threshold limit, which improves the accuracy and stability of automatic exposure of high speed camera. The experimental results show that this method can measure the brightness of a frame of image in 2ms and adjust the exposure time. Compared with the traditional automatic exposure method based on average luminance value, in 0~110ms, the illumination intensity of the light source changes from 760,23 100lux. The entropy of image information obtained by this method is 48.38 higher than that of the average brightness value method, and the variance is reduced by 62.13 points, which can provide better and more stable image detail information for subsequent automatic focusing. Image recognition and target tracking provide reference.
【作者单位】： 中国科学院长春光学精密机械与物理研究所;中国科学院大学;
【基金】：国家863高技术研究发展计划资助项目(No.2011AA8082035)
【分类号】：TP391.41
【相似文献】
相关期刊论文 前5条
1 谷明琴;蔡自兴;易亮;;面向无人驾驶车感知系统的自动曝光方法[J];计算机科学;2013年06期
2 陈参宁;邓华秋;王剑华;;基于光圈控制的自动曝光算法研究[J];传感器与微系统;2011年11期
3 李华;;基于FPGA的成像系统自动曝光算法设计[J];商洛学院学报;2014年02期
4 许世文,于洪利,王岩;CCD光电转换的自动曝光技术[J];哈尔滨工业大学学报;1991年05期
5 ;[J];;年期
相关硕士学位论文 前4条
1 仵宗钦;基于FPGA的数字相机自动曝光技术研究[D];长春理工大学;2016年
2 宁静怡;基于信息熵的自动曝光算法实现的研究[D];中国航天科技集团公司第一研究院;2016年
3 杨作廷;基于月球探测全景相机自动曝光技术的研究[D];中国科学院研究生院(西安光学精密机械研究所);2012年
4 崔阳;基于CIS芯片的自动曝光电路设计[D];天津大学;2010年
，本文编号：2421308
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2421308.html
