Using Depth to Extend Randomised Hough Forests for Object Detection and Localisation
Implicit Shape Models (ISM) have been developed for object detection and localisation in 2-D (RGB) imagery and, to a lesser extent, full 3-D point clouds. Research is ongoing to extend the approach to 2-D imagery having co-registered depth (RGB- D) e.g. from stereoscopy, laser scanning, time-of-flig...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Conference Paper |
| Published: |
IEEE Inc.
2013
|
| Online Access: | http://hdl.handle.net/20.500.11937/21871 |
| _version_ | 1848750711763894272 |
|---|---|
| author | Palmer, Richard Tan, Tele West, Geoff |
| author2 | Prof. Paulo de Souza |
| author_facet | Prof. Paulo de Souza Palmer, Richard Tan, Tele West, Geoff |
| author_sort | Palmer, Richard |
| building | Curtin Institutional Repository |
| collection | Online Access |
| description | Implicit Shape Models (ISM) have been developed for object detection and localisation in 2-D (RGB) imagery and, to a lesser extent, full 3-D point clouds. Research is ongoing to extend the approach to 2-D imagery having co-registered depth (RGB- D) e.g. from stereoscopy, laser scanning, time-of-flight cameras etc. A popular implementation of the ISM is as a Randomised Forest of classifier trees representing codebooks for use in a Hough Transform voting framework. We present three extensions to the Class-Specific Hough Forest (CSHF) that utilises RGB and co- registered depth imagery acquired via stereoscopic mobile imaging. We demonstrate how depth and RGB information can be combined during training and at detection time. Rather than encoding depth as a new dimension of Hough space (which can increase vote sparsity), depth is used to modify the resulting placement and strength of votes in the original 2-D Hough space. We compare the effect of these depth-based extensions to the unmodified CSHF detection framework evaluated against a challenging new real- world dataset of urban street scenes. |
| first_indexed | 2025-11-14T07:41:11Z |
| format | Conference Paper |
| id | curtin-20.500.11937-21871 |
| institution | Curtin University Malaysia |
| institution_category | Local University |
| last_indexed | 2025-11-14T07:41:11Z |
| publishDate | 2013 |
| publisher | IEEE Inc. |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | curtin-20.500.11937-218712023-02-08T04:04:14Z Using Depth to Extend Randomised Hough Forests for Object Detection and Localisation Palmer, Richard Tan, Tele West, Geoff Prof. Paulo de Souza Dr. Ulrich Engelke Dr. Ashfaqur Rahman Implicit Shape Models (ISM) have been developed for object detection and localisation in 2-D (RGB) imagery and, to a lesser extent, full 3-D point clouds. Research is ongoing to extend the approach to 2-D imagery having co-registered depth (RGB- D) e.g. from stereoscopy, laser scanning, time-of-flight cameras etc. A popular implementation of the ISM is as a Randomised Forest of classifier trees representing codebooks for use in a Hough Transform voting framework. We present three extensions to the Class-Specific Hough Forest (CSHF) that utilises RGB and co- registered depth imagery acquired via stereoscopic mobile imaging. We demonstrate how depth and RGB information can be combined during training and at detection time. Rather than encoding depth as a new dimension of Hough space (which can increase vote sparsity), depth is used to modify the resulting placement and strength of votes in the original 2-D Hough space. We compare the effect of these depth-based extensions to the unmodified CSHF detection framework evaluated against a challenging new real- world dataset of urban street scenes. 2013 Conference Paper http://hdl.handle.net/20.500.11937/21871 10.1109/DICTA.2013.6691536 IEEE Inc. restricted |
| spellingShingle | Palmer, Richard Tan, Tele West, Geoff Using Depth to Extend Randomised Hough Forests for Object Detection and Localisation |
| title | Using Depth to Extend Randomised Hough Forests for Object Detection and Localisation |
| title_full | Using Depth to Extend Randomised Hough Forests for Object Detection and Localisation |
| title_fullStr | Using Depth to Extend Randomised Hough Forests for Object Detection and Localisation |
| title_full_unstemmed | Using Depth to Extend Randomised Hough Forests for Object Detection and Localisation |
| title_short | Using Depth to Extend Randomised Hough Forests for Object Detection and Localisation |
| title_sort | using depth to extend randomised hough forests for object detection and localisation |
| url | http://hdl.handle.net/20.500.11937/21871 |