Flight Pattern Generation for Automated Aerial Imaging of Crops

Abstract

Small unmanned aerial vehicles (sUAVs) equipped with multispectral imaging system provide means for quick, low-cost, autonomous, and high resolution imaging that could support the crop management system for onions and peaches. Typically, these crops are grown over large areas and the imaging system has a limited field of view (FOV). To address this problem, multiple images have to be captured and then stitched together to create a high resolution image that can then be processed. An algorithm was developed to generate a custom flight pattern for the sUAV taking into account the desired altitude, camera’s FOV, and field size and orientation. The generated flight pattern allows the imaging system to cover the desired region of interest. Then the multiple images are stitched into a single image and subsequent image processing and analysis are conducted to accomplish specific goals such monitoring crop health such as estimation of vegetation indices, crop population, or blossom density. Initial results show that the imaging system combined with the flight pattern algorithm has the potential to be a crop management tool.

This document is currently not available here.

Share

COinS
 

Flight Pattern Generation for Automated Aerial Imaging of Crops

Small unmanned aerial vehicles (sUAVs) equipped with multispectral imaging system provide means for quick, low-cost, autonomous, and high resolution imaging that could support the crop management system for onions and peaches. Typically, these crops are grown over large areas and the imaging system has a limited field of view (FOV). To address this problem, multiple images have to be captured and then stitched together to create a high resolution image that can then be processed. An algorithm was developed to generate a custom flight pattern for the sUAV taking into account the desired altitude, camera’s FOV, and field size and orientation. The generated flight pattern allows the imaging system to cover the desired region of interest. Then the multiple images are stitched into a single image and subsequent image processing and analysis are conducted to accomplish specific goals such monitoring crop health such as estimation of vegetation indices, crop population, or blossom density. Initial results show that the imaging system combined with the flight pattern algorithm has the potential to be a crop management tool.