Chapter 8

UAV Remote Sensing in Precision Agriculture

Digital Image Processing and Analytics of Remote-Sensed Data

UAV image processing and analytics constitute a crucial phase in harnessing the vast data captured by UAVs for insightful applications. This multifaceted process involves a series of sophisticated techniques and algorithms.

Digital Image Processing

The general image processing and analytics steps include preprocessing, segmentation, feature extraction, data analysis, and evaluation. The main goal of image preprocessing is to improve the separation between plant and non plant structures in the color space by enhancing their topological representation. Segmentation is a vital step in image processing, especially for geographic object-based analysis, directly influencing feature extraction and classification quality. It entails dividing an image into homogeneous regions based on properties like color, texture, and size, employing techniques such as edge-based, region-based, pixel-based, and hybrid-based segmentation. Deep learning-based segmentation, utilizing algorithms like CNN and PlantU-net, is increasingly popular in UAV-based image processing.

Digital Image Analysis

Geographic information systems (GIS) software performs several key functions, including data input, storage, analysis, and visualization. It allows users to create, manage, and analyze geographically referenced data, ultimately providing insights into spatial relationships and patterns. It integrates spatial data with drone imagery to produce detailed maps that can highlight variations in crop health, moisture levels, and more. The most commonly used software for data processing includes geographic information system like ENVI, ERDAS IMAGINE, ArcGIS, TNTMips, GRASS GIS, QGIS, and TerrSet.

Click on the following topics for more information on UAV remote sensing in precision agriculture.