Improved Point-Line Feature Based Visual SLAM Method for Complex Environments

Fei Zhou,Limin Zhang,Chaolong Deng,Xinyue Fan

DOI: https://doi.org/10.3390/s21134604

IF: 3.9

2021-07-05

Sensors

Abstract:Traditional visual simultaneous localization and mapping (SLAM) systems rely on point features to estimate camera trajectories. However, feature-based systems are usually not robust in complex environments such as weak textures or obvious brightness changes. To solve this problem, we used more environmental structure information by introducing line segments features and designed a monocular visual SLAM system. This system combines points and line segments to effectively make up for the shortcomings of traditional positioning based only on point features. First, ORB algorithm based on local adaptive threshold was proposed. Subsequently, we not only optimized the extracted line features, but also added a screening step before the traditional descriptor matching to combine the point features matching results with the line features matching. Finally, the weighting idea was introduced. When constructing the optimized cost function, we allocated weights reasonably according to the richness and dispersion of features. Our evaluation on publicly available datasets demonstrated that the improved point-line feature method is competitive with the state-of-the-art methods. In addition, the trajectory graph significantly reduced drift and loss, which proves that our system increases the robustness of SLAM.

engineering, electrical & electronic,chemistry, analytical,instruments & instrumentation

What problem does this paper attempt to address?

The paper primarily addresses the robustness issues of Visual Simultaneous Localization and Mapping (VSLAM) systems in complex environments. Traditional point feature-based VSLAM systems perform poorly in environments with weak textures or significant lighting changes. To solve this problem, the researchers proposed an improved visual SLAM method that integrates point and line features.

Specifically, the main contributions of the paper include:

1. **Proposed an adaptive ORB algorithm**: The algorithm adjusts the threshold adaptively based on the brightness of image blocks, making the distribution of extracted features more uniform and improving the algorithm's robustness to sudden lighting changes.

2. **Improved the traditional line feature processing method**: By utilizing the results of point feature matching to enhance the accuracy of line feature matching, and introducing methods for merging line segments and deleting short lines to reduce processing time and improve matching accuracy.

3. **Proposed a new reprojection error cost function**: The function weights the reprojection errors of point and line features based on the richness and dispersion of features in the scene, fully leveraging the advantages of both point and line features.

In summary, this paper aims to improve the localization accuracy and stability of VSLAM systems in complex environments by integrating information from both point and line features.