介绍了一种可用于小型水下机器人的前视声纳信息提取方法.利用该方法获取了声纳视域内目标的方位信息,这在小型水下机器人的自主目标跟踪和避碰方面具有很大的应用价值.该方法主要包括3个部分的内容:利用声纳数据生成声纳图像；对声纳图像进行预处理；从处理后的图像中提取目标的特征信息.针对图像中较大目标的边缘信息,提出了一种基于最小二乘法的分段曲线拟合的方法,并给出了基于实验室的水池中获得的实测数据的拟合结果,验证了该方法的有效性.%A method of extracting information from raw sonar scanline data generated by a single-beam FLS mounted on a small UUV (unmanned underwater vehicle) is described. The bearing and distance information has been acquired through this method and this result could be highly used in the problems of auto-tracking and obstacle detection and avoidance. This procedure includes generating sonar image by raw sonar scanline data, processing the raw sonar image and extracting feature information of objects detected in the processed image . Aiming at extracting the edge feature information of a large object, a segmented curve fitting method is proposed based on least squares theory and the effectiveness of the method is verified by the experiment of processing the actual data acquired from the laboratory pool.
展开▼
