Fuzzy logic control of a multispectral imaging sensor for in-field plant sensing

  • Authors:

  • Yunseop Kim;John F. Reid;Qin Zhang

  • Affiliations:

  • Northern Plains Agricultural Research Laboratory, USDA-ARS, 1500 North Central Avenue, Sidney, MT 59270, USA;Moline Technology Innovation Center, John Deere, One John Deere Place, Moline, IL 61265, USA;Agricultural and Biological Engineering, University of Illinois at Urbana-Champaign 1304 W. Pennsylvania Avenue, Urbana, IL 61801, USA

  • Venue:
  • Computers and Electronics in Agriculture
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

The development of an in-field plant sensing system for a site-specific application can protect the environment from excessive chemicals and save management cost while maintaining productivity. A multi-spectral imaging sensor has been introduced and widely used for in-field plant sensing. In order for a robust performance of the spectral imaging sensor under changes in ambient illumination, image quality must be maintained for proper spectral image analysis. Image formation that is affected by camera parameters was identified, and a controller was developed to compensate varying image intensity and to obtain the desired image quality. A fuzzy logic control algorithm was applied to automatically adjust the camera exposure and gain to control image brightness within a targeted gray level. Slow convergence and oscillation were regulated by dynamic membership functions with different weights in each image channel. Images affected by illumination disturbance quickly converged into a desired brightness image within a maximum of five iterations over the entire range of camera gains in all three spectral image channels. An application of in-field plant sensing using the fuzzy logic image controller was evaluated on corn crops for nitrogen detection. The normalized spectral response of the sensor was inversely correlated to a chlorophyll meter with -0.93 and -0.88 in red and green channels, respectively. The development of an image quality controller using fuzzy logic enhanced the reliable performance of the in-field plant sensing system.