A study of spectral indices and visual characteristics for green coverage classification in aerial images of crops

• Grant number: 17/19928-4
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: November 01, 2017
• End date: September 30, 2019
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Moacir Antonelli Ponti
• Grantee: Tobias Mesquita Silva da Veiga
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

Nowadays, agriculture is highly competitive, demanding more precise and advanced technology such as remote sensing. Due to the development of such tools, there is a higher volume of data available, and consequently the need to use better methods to analyze such data in order to make it useful to farmers. In this project, we study the performance of image processing and machine learning methods in datasets of low altitude remore sensing images, acquired under different conditions. The images are taken from controled crop areas allowing the study of different scenarios. We propose the evaluation of methods for feature extraction such as vegetation indices, texture, color histogram, as well as convolutional neural network-based features. Those features are going to be used as input for supervised learning algorithms, in particular those based on Boosting. As a result, we expect to conclude which set of features are more adequate to represent the images in the task at hand. (AU)