README.md

Welcome to My Science Board

This board is recording my research progress of underwater SLAM, including papers I read, projects I run and plans I made.

Papers

I think there are three types of papers. Following papers are all about SLAM.

1. Review

• Hidalgo, Franco, 和Thomas Bräunl. 《Review of underwater SLAM techniques》. 收入 2015 6th International Conference on Automation, Robotics and Applications (ICARA), 306–311. IEEE, 2015.||16
• Cadena, Cesar, Luca Carlone, Henry Carrillo, Yasir Latif, Davide Scaramuzza, José Neira, Ian Reid和John J. Leonard. 《Past, present, and future of simultaneous localization and mapping: Toward the robust-perception age》. IEEE Transactions on robotics 32, 期 6 (2016年): 1309–1332.||1438

2. Method

• 施小成, 和王晓娟. 《一种面向AUV水下对接的双目视觉测距方法》. 计算机测量与控制, 期 10 (2008年): 1460-1462+1488.
• 张勋, 肖遥和李凡贡. 《基于测距声纳与光视觉的水下目标定位方法研究》. 船舶工程 38, 期 05 (2016年): 74–78.
• Mallios, Angelos, Pere Ridao, David Ribas和Emili Hernández. 《Scan matching SLAM in underwater environments》. Autonomous Robots 36, 期 3 (2014年): 181–198.||63
• Roznere, Monika, 和Alberto Quattrini Li. 《Underwater Monocular Image Depth Estimation using Single-beam Echosounder》, 不详.
• Zhang, Kaixiang, Jian Chen和Bingxi Jia. 《Asymptotic Moving Object Tracking with Trajectory Tracking Extension: A Homography-Based Approach: HOMOGRAPHY-BASED MOVING OBJECT TRACKING》. International Journal of Robust and Nonlinear Control 27, 期 18 (2017年12月1日): 4664–85.

3. Application

Projects

I will run some SLAM demo on Github. In order to realize docking, I will create my own system.

1. SLAM

2. Object Detection

2020-11-29 A BERT model for classify style of 5 writers'(LuXun,WangXiaobo,QianZhongshu,Moyan,ZhangAiling) article. 97% accuracy.

Plans

Before 2021, I think 50 abstracts are needed at least.
In addition, I should design a method to dock efficiently.

Ref

Datasets:UDID
KITTI etc.

Tools

• Roam Research
• Zotero
• 印象笔记

ThisSite

Main

一种面向AUV水下对接的双目视觉测距方法

提出了改进的NCC方法，采用了图像金字塔、纹理控制和视差控制改进了传统NCC方法的时间性能，双向匹配保证精度。模拟实验证明有一定效果，但未进行水试。
可学习的点主要有对接驳的特性应用、图像金字塔、算法性能的实用分析对比。
返回board

基于测距声纳与光视觉的水下目标定位方法研究

用两个声纳和一个光学相机完成测距。通过测距声纳计算得到目标中心的位置，然后在视觉图像上标记出来并提取边缘角点，最后算出目标长宽以及中心位置。
可学习的点主要有传统图像处理(顶帽变换)、传感器布设与结合、几何分析。
返回board

Review of underwater SLAM techniques

This paper presents a review of the approaches used in state-of-the-art SLAM techniques:Extented Kalman Filters SLAM(EKF-SLAM), FastSLAM, GraphSLAM and its application in underwater environments.
返回board

Past, present, and future of simultaneous localization and mapping: Toward the robust-perception age

We survey the current state of SLAM and consider future directions.
返回board

Scan matching SLAM in underwater environments

This paper proposes a pose-based algorithm to solve the full simutaneously localization and mapping problem for AUVs in unknown and possibaly unstructured environments.
返回board

Underwater Monocular Image Depth Estimation using Single-beam Echosounder

论文的主要贡献在于提供了标定声纳测深仪和相机的标定方法，并且用于校正单目相机的深度估计，这一方法还能用于颜色校正。
可学习的点主要在于相机和深度仪融合的方式：最小化测量值与估计值的误差来获得位姿，再计算校正后的深度。
返回board

Handbook of Marine Hydrodynamics and Motion Control

包含涉水载具的各种模型，运动学、流体静力学、流体动力学等。 返回board

Asymptotic Moving Object Tracking with Trajectory Tracking Extension: A Homography-Based Approach: HOMOGRAPHY-BASED MOVING OBJECT TRACKING

设计了一个基于单应矩阵的控制器，然后设计了开环和闭环误差并利用RISE反馈框架输出了相应控制信号。 返回board