Combining Deep Learning and Robust Estimation for Outlier-Resilient Underwater Visual Graph SLAM

Abstract

Visual Loop Detection (VLD) is a core component of any Visual Simultaneous Localization and Mapping (SLAM) system, and its goal is to determine if the robot has returned to a previously visited region by comparing images obtained at different time steps. This paper presents a new approach to visual Graph-SLAM for underwater robots that goes one step forward the current techniques. The proposal, which centers its attention on designing a robust VLD algorithm aimed at reducing the amount of false loops that enter into the pose graph optimizer, operates in three steps. In the first step, an easily trainable Neural Network performs a fast selection of image pairs that are likely to close loops. The second step carefully confirms or rejects these candidate loops by means of a robust image matcher. During the third step, all the loops accepted in the second step are subject to a geometric consistency verification process, being rejected those that do not fit with it. The accepted loops are then used to feed a Graph-SLAM algorithm. The advantages of this approach are twofold. First, the robustness in front of wrong loop detection. Second, the computational efficiency since each step operates only on the loops accepted in the previous one. This makes online usage of this VLD algorithm possible. Results of experiments with semi-synthetic data and real data obtained with an autonomous robot in several marine resorts of the Balearic Islands, support the validity and suitability of the approach to be applied in further field campaigns.

A Review of Underwater Robot Localization in Confined Spaces

“…T. Manderson et al [83] applied deep learning visual SLAM to underwater collision avoidance for AUVs, and M. Leonardi et al [84] used deep learning in underwater image enhancement, removing unimportant feature points from the image to leave high-quality points, thus accelerating the operation speed of underwater visual SLAM. A significant amount of work using deep learning visual SLAM has focused on loop closure detection [85][86][87][88]. This is because geometry-based visual SLAM performs poorly in the loop closure detection stage, while deep learning-based image detection technology is mature and performs excellently, greatly improving the success rate of loop closure detection.…”

Current study on multi-robot collaborative vision SLAM

“…In document [19], it provides an improved vision-based PG-SLAM graph optimization algorithm to reduce the large amount of computation caused by false positives when optimizing the graph. The principle is that the Siamese convolutional neural network (SCNN) compares two images and outputs existing overlapping information about the two parts of the environment they describe, which is verified by image-based cyclic filtering (ILF).…”

Group-$k$ Consistent Measurement Set Maximization for Robust Outlier Detection

“…Work by Burguera et al (2022) introduces a three step process that combines deep learning with RANSAC and a geometric verification step. They utilize a neural network to detect possible matches between images.…”