Summary of Evaluation Of Deep Learning Semantic Segmentation For Land Cover Mapping on Multispectral, Hyperspectral and High Spatial Aerial Imagery, by Ilham Adi Panuntun et al.
Evaluation of Deep Learning Semantic Segmentation for Land Cover Mapping on Multispectral, Hyperspectral and High Spatial Aerial Imagery
by Ilham Adi Panuntun, Ying-Nong Chen, Ilham Jamaluddin, Thi Linh Chi Tran
First submitted to arxiv on: 20 Jun 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed study explores deep learning semantic segmentation techniques for accurate land cover mapping using multiple image types, including multispectral, hyperspectral, and high spatial aerial images. The authors implemented various models such as Unet, Linknet, FPN, and PSPnet to categorize vegetation, water, and other classes (soil and impervious surface). Notably, the LinkNet model achieved high accuracy with an Intersection Over Union (IoU) of 0.92 across all datasets, comparable to other techniques. The study found that multispectral images showed higher performance in terms of IoU and F1-score, highlighting their potential for land cover classification. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Land cover mapping is crucial for environmental monitoring, especially with climate change. This study uses deep learning to improve accuracy and speed up the process. They tested different models on various types of satellite images, including multispectral, hyperspectral, and high spatial aerial images. The best model, called LinkNet, was very accurate, beating other methods in some cases. The results show that using multispectral images can give even better results. |
Keywords
» Artificial intelligence » Classification » Deep learning » F1 score » Semantic segmentation » Unet
