激光水下成像系统辐射标定算法  

A radiometric calibration algorithm for an underwater laser-imaging system

作　　者：肖易寒[1] 庞永杰[2] 赵蓝飞[1] 

机构地区：[1]哈尔滨工程大学信息与通信工程学院,黑龙江哈尔滨150001 [2]哈尔滨工程大学水下机器人技术重点实验室,黑龙江哈尔滨150001

出　　处：《哈尔滨工程大学学报》2016年第5期738-742,共5页Journal of Harbin Engineering University

基　　金：国家自然科学基金资助项目(51179035/51279221);中央高校基本科研业务费专项资金资助(HEUCF160802)

摘　　要：由于像增强器的非线性响应破坏了散斑噪声的统计规律,因此直接对激光水下图像进行散斑噪声抑制得到的图像复原结果受到非线性响应的约束。为了恢复散斑噪声的固有特性,本文提出一种对光晕具有鲁棒性的辐射标定算法。这种辐射标定算法通过灰阶映射函数以及积分时间增量,将非线性响应曲线的非线性部分转化为线性部分,再通过线性插值标定出像增强器对于激光信号的非线性响应关系,从而达到恢复散斑噪声的分布规律,提高噪声抑制效果的目的。通过实验对比了辐射标定前、后激光水下图像的噪声抑制效果,验证了本文算法能够有效地提高散斑噪声的抑制效果。Because the nonlinear response of an image intensifier influences the distribution of inherent noise, the effect of speckle noise reduction is limited through the laser＇s underwater low light level images. In order to recover the distribution of the inherent noise, this paper proposes a radiometrie calibration algorithm that is robust to halos ; the algorithm is used to calibrate the nonlinear response function of the image intensifier. This algorithm employs a gray-scale mapping function and an integral time increment to transform the nonlinear response into a linear re- sponse. Subsequently, the linear interpolation is employed to calibrate the nonlinear response of the image intensifi- er with regard to laser radiance. The distribution of inherent noise is thereby restored, which contributes to impro- ving the effect of speckle noise reduction. In the experimental section, we compare the restoration with a radiometrie calibration and with a conventional speckle-reduction algorithm. The comparison illustrates that this algorithm effectively improves the performance of speckle noise reduction.

关 键 词：辐射标定 像增强器 激光水下图像 散斑噪声 相机响应函数 光晕 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]