{"id":{"repo_id":"nus","oai_identifier":"oai:scholarbank.nus.edu.sg:10635/309812"},"canonical_url":"https://search.dev.ndltd.org/etd/nus/oai:scholarbank.nus.edu.sg:10635/309812","repository":{"repo_id":"nus","name":"National University of Singapore","base_url":"https://scholarbank.nus.edu.sg/oai/request"},"display":{"title":"MONOCULAR VISUAL ODOMETRY","abstract":"When estimating a robot's egomotion from visual input alone using methods such as visual odometry (VO) or visual simultaneous localization and mapping (V-SLAM), motion can only be recovered only up to an unknown scale factor in a monocular setup. Visual place recognition (VPR) can provide absolute pose estimates by matching the current camera frame to a database of reference images labeled with known coordinates. By using VPR to provide absolute pose estimates in the map and matching these estimates to the relative poses from monocular VO or V-SLAM, the absolute scale of the trajectory given by monocular VO or V-SLAM can be found. This thesis aims to demonstrate the application of this approach to achieve absolute pose estimation with a single camera. Results indicate that the approach corrected scale drift, improved the accuracy of the pose estimate, and yielded a globally consistent trajectory that closely follows the ground truth trajectory.","abstract_html":"When estimating a robot's egomotion from visual input alone using methods such as visual odometry (VO) or visual simultaneous localization and mapping (V-SLAM), motion can only be recovered only up to an unknown scale factor in a monocular setup. Visual place recognition (VPR) can provide absolute pose estimates by matching the current camera frame to a database of reference images labeled with known coordinates. By using VPR to provide absolute pose estimates in the map and matching these estimates to the relative poses from monocular VO or V-SLAM, the absolute scale of the trajectory given by monocular VO or V-SLAM can be found. This thesis aims to demonstrate the application of this approach to achieve absolute pose estimation with a single camera. Results indicate that the approach corrected scale drift, improved the accuracy of the pose estimate, and yielded a globally consistent trajectory that closely follows the ground truth trajectory.","abstract_has_math":false,"creators":["BEATRIX TUNG XUE LIN"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025-01-02","date_published":"2025-01-02","updated_at":"2026-07-24T03:31:51Z","subjects":["visual place recognition","visual simultaneous localization and mapping","visual odometry","monocular"],"languages":[],"rights":[],"rights_urls":["https://scholarbank.nus.edu.sg/bitstreams/23e07fb2-19b7-44e3-8b97-832e46439ffc/download"],"identifier_entries":[]},"links":{"outbound_url":null,"outbound_label":null,"outbound_source":null},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["BEATRIX TUNG XUE LIN"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.issued","label":"Date","values":["2025-01-02"]},{"key":"dc:relation.isreferencedby","label":"Dc Relation Isreferencedby","values":["https://scholarbank.nus.edu.sg/handle/10635/309812"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["visual place recognition","visual simultaneous localization and mapping","visual odometry","monocular"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["https://scholarbank.nus.edu.sg/bitstreams/23e07fb2-19b7-44e3-8b97-832e46439ffc/download"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://scholarbank.nus.edu.sg/bitstreams/f7beb1c6-4ca0-4c45-b1a2-2c46f6d323ca/download"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["When estimating a robot's egomotion from visual input alone using methods such as visual odometry (VO) or visual simultaneous localization and mapping (V-SLAM), motion can only be recovered only up to an unknown scale factor in a monocular setup. Visual place recognition (VPR) can provide absolute pose estimates by matching the current camera frame to a database of reference images labeled with known coordinates. By using VPR to provide absolute pose estimates in the map and matching these estimates to the relative poses from monocular VO or V-SLAM, the absolute scale of the trajectory given by monocular VO or V-SLAM can be found. This thesis aims to demonstrate the application of this approach to achieve absolute pose estimation with a single camera. Results indicate that the approach corrected scale drift, improved the accuracy of the pose estimate, and yielded a globally consistent trajectory that closely follows the ground truth trajectory."]},{"key":"dc:format.checksum.md5","label":"Dc Format Checksum Md5","values":["9f3c6f2aa8f96291ef77b0a18484521d","d1dc65f55233e1e9ecfdddae750c89bd","a4a47f63485052d0d8bc2765b2613db7"]},{"key":"dc:title","label":"Title","values":["MONOCULAR VISUAL ODOMETRY"]}]}],"canonical_facts":{"dc:creator":["BEATRIX TUNG XUE LIN"],"dc:date.issued":["2025-01-02"],"dc:description.abstract":["When estimating a robot's egomotion from visual input alone using methods such as visual odometry (VO) or visual simultaneous localization and mapping (V-SLAM), motion can only be recovered only up to an unknown scale factor in a monocular setup. Visual place recognition (VPR) can provide absolute pose estimates by matching the current camera frame to a database of reference images labeled with known coordinates. By using VPR to provide absolute pose estimates in the map and matching these estimates to the relative poses from monocular VO or V-SLAM, the absolute scale of the trajectory given by monocular VO or V-SLAM can be found. This thesis aims to demonstrate the application of this approach to achieve absolute pose estimation with a single camera. Results indicate that the approach corrected scale drift, improved the accuracy of the pose estimate, and yielded a globally consistent trajectory that closely follows the ground truth trajectory."],"dc:format.checksum.md5":["9f3c6f2aa8f96291ef77b0a18484521d","d1dc65f55233e1e9ecfdddae750c89bd","a4a47f63485052d0d8bc2765b2613db7"],"dc:identifier.uri":["https://scholarbank.nus.edu.sg/bitstreams/f7beb1c6-4ca0-4c45-b1a2-2c46f6d323ca/download"],"dc:relation.isreferencedby":["https://scholarbank.nus.edu.sg/handle/10635/309812"],"dc:rights":["https://scholarbank.nus.edu.sg/bitstreams/23e07fb2-19b7-44e3-8b97-832e46439ffc/download"],"dc:subject":["visual place recognition","visual simultaneous localization and mapping","visual odometry","monocular"],"dc:title":["MONOCULAR VISUAL ODOMETRY"],"dc:type":["Thesis"]},"updated_at":"2026-07-24T03:31:51Z"}