An extensive analysis on oriented object detection

Date

2023-11-30

Author

Koç, İbrahim

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

79
views

138
downloads

Cite This

Object Detection, one of the landmark problems in Computer Vision, is conventionally addressed by placing horizontal rectangular boxes around objects. The use of such bounding boxes, however, puts a restriction about the orientations of the objects. For elongated and oriented objects, such boxes usually include disruptive visual information coming from background or other objects, which can lead to poor detection performance for such elongated and oriented objects. To address this issue, in Oriented Object Detection (OOD), objects are described and detected with oriented rectangular bounding boxes. The introduction of many large-scale datasets with oriented bounding box annotations, e.g. the DOTA-v1.0 dataset, has facilitated a plethora of approaches for OOD. However, to the best of our knowledge, no study extensively analyzes the issues pertaining to the OOD task. In this thesis, we provide detailed analyzes using the DOTA-v1.0 dataset and show that (i) there is a severe imbalance problem regarding the distribution of objects across different orientations and scales, especially for certain object classes (e.g. small vehicle, bridge, harbor), and (ii) OOD models utilizing different approaches also demonstrate imbalance in terms of orientation distribution. Taking into account the outcomes of our analyses, we demonstrate enhancements in various orientations and scales within Oriented Object Detection models.

Subject Keywords

Object Detection, Oriented Object Detection, Scale and Orientation Imbalance

URI

https://hdl.handle.net/11511/107717

Collections

Graduate School of Natural and Applied Sciences, Thesis