Systematic literature review on approaches of extracting image merits
الباحث الأول:
Ameer N. Onaizah
الباحثين الآخرين:
Yuanqing Xia
Yufeng zhan
Khurram hussain
المجلة:
Optik - International Journal for Light and Electron Optics
تاريخ النشر:
17 ديسمبر، 2022
مختصر البحث:
Texture analysis is gaining popularity among the scientific community. A wide variety of applications
use texture analysis method. Texture analysis methods can be used for image segmentation,
pattern analysis and pattern classification tasks. The …
Texture analysis is gaining popularity among the scientific community. A wide variety of applications
use texture analysis method. Texture analysis methods can be used for image segmentation,
pattern analysis and pattern classification tasks. The application areas range from remote
sensing, biomedical imaging, image synthesis, image inpainting and image processing. However,
the preliminary step in all these applications refers to the extraction of intricate features from the
given image. As a result, a wide verity of feature extraction methods exists in the literature. All
the feature extraction methods have their own advantages and shortfalls. For example, some of
the methods are computationally expensive, some are rotation and scale invariant whereas the
others are easy to implement. This article provides an insight regarding different texture feature
extraction techniques. The article bifurcate these techniques into different techniques according
to their working principles. Besides provision of the basic working principle of every technique,
the article provides an insight regarding their advantages and shortfalls. Moreover, this article
considers deep learning and entropy based methods interesting for texture evaluation. Besides,
the article also proposes a thorough study of these methods in texture analysis.
