海洋激光雷达图像处理提取海水深度的方法

Extracting sea water depth by image processing of ocean lidar

机载激光雷达已经应用于浅海地形测绘,与激光器、接收望远镜和探测器一样,水深提取算法也是决定系统最大测深能力的关键环节。常规的水深提取算法是对单个激光雷达采集的波形数据进行处理,通过提取波形中的海表和海底位置实现水深测量,这种方法在提取水深较深的海底微弱回波信号时,易受海水散射层强信号的影响,导致水深提取能力和准确度下降。为了解决这一问题,将一维回波波形数据按采集顺序组合成二维的回波强度图像,图像的每一列代表一条回波波形,图像的灰度值对应着回波信号强度。利用图像的横向相关性,通过双边滤波、局部阈值二值化等图像处理方法,提取出海底回波信号廓线。该方法一方面提升了海底回波的提取能力,一方面避免散射层信号对海底微弱信号的干扰,为浅海地形、水下目标一体化探测提供新的数据处理方式。

Airborne lidars have been used in shallow sea surveying and mapping.The water depth extraction algorithm is vital in assessing the system’s maximum sounding ability,similar to lasers,receiving telescopes,and detectors.The conventional water depth extraction algorithm processes the single waveform data and measures water depth through the extraction of the sea surface and seafloor positions in the waveform.This method is susceptible to the influence of strong signals from the seawater scattering layer when extracting weak echo signals from the bottom of the deep water,resulting in a decrease in the ability and accuracy of water depth extraction.To address this issue,the one-dimensional echo waveform data was combined into two-dimensional echo intensity images according to the acquisition order.Each column of the image represented an echo waveform,and the gray value of the image corresponded to the intensity of the echo signal.Using the horizontal correlation of the image and through image processing methods such as bilateral filtering and local threshold binarization,the submarine echo signal profile was extracted.This method both improves the extraction ability of seabed echoes and circumvents the interference of scattering layer signals on weak seabed signals while also providing a new data processing method for the combined detection of shallow sea terrain and underwater targets.