Denosing is one of an essential step to improve the image quality. In this project, image denoising is investigated. After reviewing standard image denoising methods as applied in the spatial, frequency and wavelet domains of the noisy image, the project embarks on the endeavor of developing and experimenting with new image denoising methods based wavelet transforms. Image denoising involves the manipulation of the image data to produce a visually high quality image. Finding efficient image denoising methods is still valid challenge in image processing. Wavelet denoising is an attempts to remove the noise present in the image while preserving the image characteristics, regardless of its frequency content. This project is intended to serve as an introduction to Wavelet processing through a set of Matlab experiments. These experiments will give an overview of three fundamental tasks in signal and image processing: approximation, denoising and compression.
Every day, satellites capture huge volumes of data and monitor the earth's surface. This is a timeconsuming process to manually classify the images into their appropriate classes. A classification system that automatically classifies the images into the proper classes is required. In this article, we have used handcrafted features and deep learning features to classify the satellite images. As handcrafted features, HLAC features have been used and for deep learning features ResNet-50 and ResNet-101 models have been used. This article provides a method for detecting HLAC features in multi-class satellite images and automatically classifying them into their appropriate classes. To classify an image, the image's features are first extracted, and then classification is performed utilizing those features. An SVM classifier is used to classify the images. The Sentinel-2 satellite images from the EuroSAT dataset were used. The accuracy and F1-score were used to assess the performance of the classifier. The SVM classifier with HLAC feature extraction had an F1-score of 86.26 and an accuracy of 89.35%. The dataset's classification accuracy was also assessed using a deep learning model. We obtained a 98.69% accuracy using deep learning models which is 0.12% higher than the benchmarked.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.