Semi-Direct Monocular SLAM With Three Levels of Parallel Optimizations

Abstract

In practical applications, how to use the complementary strengths of the direct and the feature-based methods for effective fusion may be the main challenge of simultaneous localization and mapping (SLAM). To solve this challenge, we propose the DO-SLAM, a novel fast and accurate semidirect visual SLAM framework, which can maintain the direct method's fast performance and the high precision and loop closure capability of the feature-based method. The direct method is used as the first half of the DO-SLAM to track the camera pose rapidly and robustly. The feature-based method is used as the second half of the DO-SLAM to refine the keyframe poses, perform loop closures, and build a globally consistent, long-term, sparse feature map that can be reused. The proposed pipeline fuses direct odometry and feature-based SLAM to perform three levels of parallel optimizations: (1) In the direct method module, the keyframe poses are estimated by minimizing the photometric error, (2) In the feature-based module, using the poses calculated by the inter-frame matching to correct and fuse the poses calculated by the direct method module as the initial poses, and the initial poses are optimized by the motion-only bundle adjustment, and (3) A pose graph optimization is used to achieve global map consistency in the presence of loop closures. Experimental evaluation on two benchmark datasets demonstrates that the proposed approach achieves higher accuracy and robustness on motion estimation compared to the other state-of-the-art methods.INDEX TERMS simultaneous localization and mapping(SLAM), semi-direct SLAM, three levels of parallel optimizations

Localization and Mapping for Self-Driving Vehicles: A Survey

“…[155] attempted to imitate the human vision character by introducing a Salient Bundle Adjustment (SBA). DO-SLAM was proposed by Lu et al [156]. They tried to combine direct-based and features-based methods to create an accurate SLAM framework.…”

A Novel Ranging and IMU-Based Method for Relative Positioning of Two-MAV Formation in GNSS-Denied Environments

“…The common methods for achieving relative positioning in GNSS-denied environments can be classified into the following three categories: visual positioning [7][8][9][10][11][12][13][14][15][16][17], inertial navigation positioning [18][19][20][21][22], and radio positioning [23,24]. The development status of each category of positioning technology is systematically discussed below, and the advantages and limitations of each technology are analyzed, providing good inspiration for the study of relative positioning methods for MAVs in GNSS-denied environments.…”

“…The visual localization methods can be broadly categorized into map-based localization [7][8][9][10][11] and map-free localization [12][13][14][15][16][17], depending on whether prior visual maps are utilized. (Note that visual refers to visual information, which is typically collected using sensors such as cameras or laser scanners.…”

“…In contrast, map-free visual localization methods do not rely on prior visual maps and instead estimate the pose of the object and surrounding environment. This category can be further divided into Visual Simultaneous Localization and Mapping (VSLAM) [12][13][14][15] and Structure from Motion (SFM) [16,17]. The VSLAM is designed for real-time processing, making it well-suited for applications such as robotics and autonomous vehicles, while SFM prioritizes accuracy and is more appropriate for offline processing applications such as digital reconstruction of scenes [16].…”

“…Reference [13] proposed a weightoptimized particle filter-based algorithm for monocular visual SLAM, which aimed to improve the slow environmental interference repair speed of traditional filtering SLAM algorithms. Three-level parallel optimization was adopted in [14], including the direct method, feature-based method, and pose graph optimization. The method in [15] involves two stages: the first stage implements a local SLAM process based on filtering techniques, while the second stage utilizes optimization-based techniques for constructing and maintaining a consistent global map of the environment, which includes addressing the loop closure problem.…”