A Toolbox for Aerial Image Acquisition and its Application to Precision Agriculture *

Abstract

Precision farming has increasingly focused on lowering production costs. Remote sensing techniques, mainly aerial imagery in the visible and infrared bands, have been employed to achieve this goal. The GPS (Global Positioning System) system allows easy geo-referencing of these images, making it possible to produce maps showing problems in the crop that can be easily located and corrected. The main problem is the high cost associated with the acquisition of aerial photographs with the necessary periodicity. This paper introduces TIA, a Toolbox for aerial Image Acquisition. TIA adds tools to manned and unmanned aircraft, easing tasks associated with acquisition and processing of aerial imagery. The first step towards implementation of TIA was a requirements analysis that produced a list of useful functions. These functions include mission planning, automatic mission execution, pilot guidance and geo-referencing of photographs and video frames using a GPS receiver. TIA was implemented in an architecture composed of three computer modules: a Palmtop computer that acts as a display/keyboard unit, a main computer and a camera controller. Each computer module has a corresponding software module. The toolbox has been tested onboard a ultralight aircraft and is currently being integrated into a fixed-wing unmanned aerial vehicle (UAV). To evaluate the usability of TIA, a test mission was carried out at a big farm, in the central savannas of Brazil. About 360 photographs were taken from an 800 ha crop area. Images were segmented using a technique based on neural networks. Results have revealed several problems on the fields, including nematode and weed infestation, irregular seeding, and water erosion. It is expected that its use in the next season at the farm will result in substantial gains in productivity, through the periodical analysis of the aerial imagery collected and the adoption of intra season corrective measures. Further work on TIA includes new tests using a near infrared camera, a stability augmentation system for light aircraft flying, an altitude laser sensor for photographic scale correction and a self-levering camera support.

Cite this paper

@inproceedings{Jnior2012ATF, title={A Toolbox for Aerial Image Acquisition and its Application to Precision Agriculture *}, author={Onofre Trindade J{\'u}nior and Rosane Minghim}, year={2012} }