Marine Debris Dataset for Forward-Looking Sonar Semantic Segmentation

本数据集名为‘Marine Debris Dataset for Forward-Looking Sonar Semantic Segmentation’，由德国人工智能研究中心创建，旨在通过前视声纳图像实现海洋垃圾的精确分割。数据集包含1868张使用ARIS Explorer 3000传感器捕获的图像，涵盖11个对象类别及背景。创建过程中，数据在人工水箱中收集，确保图像质量。该数据集主要应用于海洋清洁机器人，以提高对海洋垃圾的检测和分割能力，解决海洋污染问题。

This dataset is named "Marine Debris Dataset for Forward-Looking Sonar Semantic Segmentation", which was developed by the German Research Center for Artificial Intelligence (DFKI). It is designed to achieve accurate semantic segmentation of marine debris using forward-looking sonar images. The dataset comprises 1,868 images captured with the ARIS Explorer 3000 sensor, covering 11 object categories and the background class. During the dataset creation process, all data was collected in an artificial water tank to ensure consistent image quality. This dataset is primarily utilized for marine cleaning robots, aiming to improve the detection and segmentation performance of marine debris and tackle marine pollution problems.