Remote sensing imagery segmentation in object-based analysis: A review of methods, optimization, and quality evaluation over the past 20 years

• Type: Another Publication in an International Scientific Journal
• Date: 2023

Authors

Ez zahouani, B (Author) Other; Ana Teodoro (Author) FCUP; El Kharki, O (Author) Other; Jianhua, L (Author) Other; Kotaridis, I (Author) Other; Yuan, XH (Author) Other; Ma, L (Author) Other

Journal

Title: Remote Sensing Applications: Society and Environment

Vol. 32

Publisher: Elsevier

Indexing

ISI Web of Knowledge - 0 Citations

Scopus - 0 Citations

Other information

Authenticus ID: P-00Y-V0A

DOI: 10.1016/j.rsase.2023.101031

Abstract (EN): Object-based image analysis (OBIA) has become a key research topic for decades and represents an attractive paradigm leading to accurate features classification and recognition. It is based on a key fundamental technical support, named segmentation, which produces physically and semantically image objects with homogeneous pixels, and it can be used in whole or in part for research related to object-oriented analysis. This paper reviews all the steps comprising segmentation and the methods for each step by reviewing literature published in relevant remote sensing journals between 2000 and 2022. In addition, segmentation methods are discussed, compared, and categorized into general and detailed categories. The SOTA researches focusing on the most advanced methods at all steps are also presented while highlighting future research opportunities and needs. The novelty of the review is the adoption of a new classification of segmentation methods by time scale, which reveal the evolution history and trend based on their internal relations. Particularly, this work covers all the approaches and methods that make up the segmentation steps, with analyses, comparisons, and future research perspectives. Finally, conclusions about the SOTA methods, critical conclusions about open challenges, and directions and recommendations for future research are presented.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 30