Primitives extraction for structural pattern recognition
This dissertation presents the investigations conducted on the various corner detection operators in terms of their working principles and detection algorithms. These operators are broadly classified under two categories: ones that require pre-processing of the images of the objects which are in bin...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2008
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/13165 |
| _version_ | 1826124446496194560 |
|---|---|
| author | Koo, Chee Keong. |
| author2 | Chan, Kap Luk |
| author_facet | Chan, Kap Luk Koo, Chee Keong. |
| author_sort | Koo, Chee Keong. |
| collection | NTU |
| description | This dissertation presents the investigations conducted on the various corner detection operators in terms of their working principles and detection algorithms. These operators are broadly classified under two categories: ones that require pre-processing of the images of the objects which are in binary format and ones that are performed directly on the images of objects in grey level without requiring the additional pre-processing steps. In view of these, the pre-processing operations such as noise reductions, edge detection and edge thinning are investigated as well. This is to provide a reference point for the comparison on the effect of these pre-processing operations on the comer operators. The results are then compared with the corner operators that do not require these pre-processing. Images of objects with different sizes, texture, shapes and under various background conditions are used in the investigations. This is to provide a comparison platform in terms of the response of these comers operators towards different contrast, shape and noise. |
| first_indexed | 2024-10-01T06:20:46Z |
| format | Thesis |
| id | ntu-10356/13165 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T06:20:46Z |
| publishDate | 2008 |
| record_format | dspace |
| spelling | ntu-10356/131652023-07-04T15:46:07Z Primitives extraction for structural pattern recognition Koo, Chee Keong. Chan, Kap Luk School of Electrical and Electronic Engineering DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Signal processing This dissertation presents the investigations conducted on the various corner detection operators in terms of their working principles and detection algorithms. These operators are broadly classified under two categories: ones that require pre-processing of the images of the objects which are in binary format and ones that are performed directly on the images of objects in grey level without requiring the additional pre-processing steps. In view of these, the pre-processing operations such as noise reductions, edge detection and edge thinning are investigated as well. This is to provide a reference point for the comparison on the effect of these pre-processing operations on the comer operators. The results are then compared with the corner operators that do not require these pre-processing. Images of objects with different sizes, texture, shapes and under various background conditions are used in the investigations. This is to provide a comparison platform in terms of the response of these comers operators towards different contrast, shape and noise. Master of Science (Computer Control and Automation) 2008-10-20T07:16:57Z 2008-10-20T07:16:57Z 1999 1999 Thesis http://hdl.handle.net/10356/13165 en 80 p. application/pdf |
| spellingShingle | DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Signal processing Koo, Chee Keong. Primitives extraction for structural pattern recognition |
| title | Primitives extraction for structural pattern recognition |
| title_full | Primitives extraction for structural pattern recognition |
| title_fullStr | Primitives extraction for structural pattern recognition |
| title_full_unstemmed | Primitives extraction for structural pattern recognition |
| title_short | Primitives extraction for structural pattern recognition |
| title_sort | primitives extraction for structural pattern recognition |
| topic | DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Signal processing |
| url | http://hdl.handle.net/10356/13165 |
| work_keys_str_mv | AT koocheekeong primitivesextractionforstructuralpatternrecognition |