A Multilayer Perceptron-Based Spherical Visual Compass Using Global Features
This paper presents a visual compass method utilizing global features, specifically spherical moments. One of the primary challenges faced by photometric methods employing global features is the variation in the image caused by the appearance and disappearance of regions within the camera’s field of...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-03-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/24/7/2246 |
| _version_ | 1827286535864582144 |
|---|---|
| author | Yao Du Carlos Mateo Omar Tahri |
| author_facet | Yao Du Carlos Mateo Omar Tahri |
| author_sort | Yao Du |
| collection | DOAJ |
| description | This paper presents a visual compass method utilizing global features, specifically spherical moments. One of the primary challenges faced by photometric methods employing global features is the variation in the image caused by the appearance and disappearance of regions within the camera’s field of view as it moves. Additionally, modeling the impact of translational motion on the values of global features poses a significant challenge, as it is dependent on scene depths, particularly for non-planar scenes. To address these issues, this paper combines the utilization of image masks to mitigate abrupt changes in global feature values and the application of neural networks to tackle the modeling challenge posed by translational motion. By employing masks at various locations within the image, multiple estimations of rotation corresponding to the motion of each selected region can be obtained. Our contribution lies in offering a rapid method for implementing numerous masks on the image with real-time inference speed, rendering it suitable for embedded robot applications. Extensive experiments have been conducted on both real-world and synthetic datasets generated using Blender. The results obtained validate the accuracy, robustness, and real-time performance of the proposed method compared to a state-of-the-art method. |
| first_indexed | 2024-04-24T10:35:07Z |
| format | Article |
| id | doaj.art-e960649b986b425397cdfce2f609c1fb |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-24T10:35:07Z |
| publishDate | 2024-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-e960649b986b425397cdfce2f609c1fb2024-04-12T13:26:35ZengMDPI AGSensors1424-82202024-03-01247224610.3390/s24072246A Multilayer Perceptron-Based Spherical Visual Compass Using Global FeaturesYao Du0Carlos Mateo1Omar Tahri2Université Bourgogne, 21000 Dijon, FranceICB UMR CNRS 6303, Université Bourgogne, 21000 Dijon, FranceICB UMR CNRS 6303, Université Bourgogne, 21000 Dijon, FranceThis paper presents a visual compass method utilizing global features, specifically spherical moments. One of the primary challenges faced by photometric methods employing global features is the variation in the image caused by the appearance and disappearance of regions within the camera’s field of view as it moves. Additionally, modeling the impact of translational motion on the values of global features poses a significant challenge, as it is dependent on scene depths, particularly for non-planar scenes. To address these issues, this paper combines the utilization of image masks to mitigate abrupt changes in global feature values and the application of neural networks to tackle the modeling challenge posed by translational motion. By employing masks at various locations within the image, multiple estimations of rotation corresponding to the motion of each selected region can be obtained. Our contribution lies in offering a rapid method for implementing numerous masks on the image with real-time inference speed, rendering it suitable for embedded robot applications. Extensive experiments have been conducted on both real-world and synthetic datasets generated using Blender. The results obtained validate the accuracy, robustness, and real-time performance of the proposed method compared to a state-of-the-art method.https://www.mdpi.com/1424-8220/24/7/2246omnidirectional camerasrobotsmachine learningglobal feature extractionrobot vision systemslocalization |
| spellingShingle | Yao Du Carlos Mateo Omar Tahri A Multilayer Perceptron-Based Spherical Visual Compass Using Global Features Sensors omnidirectional cameras robots machine learning global feature extraction robot vision systems localization |
| title | A Multilayer Perceptron-Based Spherical Visual Compass Using Global Features |
| title_full | A Multilayer Perceptron-Based Spherical Visual Compass Using Global Features |
| title_fullStr | A Multilayer Perceptron-Based Spherical Visual Compass Using Global Features |
| title_full_unstemmed | A Multilayer Perceptron-Based Spherical Visual Compass Using Global Features |
| title_short | A Multilayer Perceptron-Based Spherical Visual Compass Using Global Features |
| title_sort | multilayer perceptron based spherical visual compass using global features |
| topic | omnidirectional cameras robots machine learning global feature extraction robot vision systems localization |
| url | https://www.mdpi.com/1424-8220/24/7/2246 |
| work_keys_str_mv | AT yaodu amultilayerperceptronbasedsphericalvisualcompassusingglobalfeatures AT carlosmateo amultilayerperceptronbasedsphericalvisualcompassusingglobalfeatures AT omartahri amultilayerperceptronbasedsphericalvisualcompassusingglobalfeatures AT yaodu multilayerperceptronbasedsphericalvisualcompassusingglobalfeatures AT carlosmateo multilayerperceptronbasedsphericalvisualcompassusingglobalfeatures AT omartahri multilayerperceptronbasedsphericalvisualcompassusingglobalfeatures |