|Alarcon, Vladimir - MISS. STATE UNIV.|
Submitted to: ASAE Annual International Meeting
Publication Type: Proceedings
Publication Acceptance Date: July 27, 2001
Publication Date: August 4, 2001
Citation: THOMSON, S.J., ALARCON, V. GEO-REFERENCED DIGITAL IMAGING AND ULTRASONIC ALTITUDE SENSING FOR AGRICULTURAL AIRCRAFT -- PILOT CONTROLS, INSTRUMENTATION, AND SYSTEM EVALUATIONS. ASAE ANNUAL INTERNATIONAL MEETING. 2002. Paper No. 013098. Interpretive Summary: Agricultural aircraft are customarily used for field spraying to control crop pests. The potential for using these aircraft to obtain field images for targeting spray is being investigated. This can save chemical and minimize spray drift because only areas that require treatment would be sprayed from the air or from ground sprayers. To be successful, pilot controls and instruments for these imaging systems must be simple to use with minimal pilot intervention required. Pilot controls, instrumentation, and three imaging systems are described. A case study in weed detection is presented, and methods used to successfully differentiate weeds from crop are described. Another remote sensing system for agricultural aircraft is being developed to determine height-of-flight for spray drift research. A system like this is necessary because spray drift is influenced by the height of spray release (e.g. from 1.5 to 9m). Preliminary evaluations using an inexpensive commercial ultrasonic sensor on agricultural aircraft are presented, along with a discussion of future research requirements.
Technical Abstract: Instrumentation are being developed for two separate remote sensing systems used on agricultural aircraft. One project was initiated to develop an imaging system for site-specific crop management. Instrumentation, pilot controls, and a digital video-based remote sensing system with geo-referencing have been adapted for convenient use by the pilot and to permit efficient image post-processing. Weeds were successfully discriminated and distinguished from cotton using classification methods in the ENVI 3.4 image analysis system. The parallelepilped and Mahalanobis distance methods allowed good discrimination of weeds and cotton with optimized distance parameters. For a second remote sensing application, preliminary data are presented for an ultrasonic altitude sensor, to be used in aerial spray application research. Preliminary data indicated a strong, repeatable signal when flying over a road surface but a questionable signal over bare soil and low crop canopy. Further testing of this method will be accomplished using a more flexible sonar-based system that allows electronic control of system parameters.