Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Weed detection in row crops using the red near-infrared reflectance ratio and frequency transforms of video images
Abstract
The use of post-emergence herbicide allows weed populations to be measured before and during the herbicide application. Since weeds are generally non-uniform in population density, herbicide savings can be realized by selective herbicide application. Such a system would estimate the local weed density, and activate the sprayer to apply herbicide to only those areas with enough weeds to justify the application. A reflectance ratio meter (RRM) and machine vision systems using color and monochrome video cameras are evaluated and compared as sensors to estimate weed population in row crops. An experiment was conducted to evaluate the accuracy of the sensors. The Fast Fourier and Fast Hadamard transforms are described and compared as machine vision and image processing techniques. The reflectance ratio meter, an optical device for detecting weeds by measuring the ratio of reflected red and near-infrared light, is described. Several methods for interpreting the signal are evaluated. The Fast Hadamard transform was shown to be as effective as the Fast Fourier transform for detecting weeds, and was over 20 times faster. Either transform could be used to classify weed and non-weed images with 90% accuracy. The near-infrared video system was superior to the color system because the classification threshold varied less by data set. The signal from the RRM was efficiently analyzed by a method approximating the signal derivative. The result correlated to the weed population with a coefficient of determination of 0.80 to 0.90 for the data on any single day.
Subject Area
Agricultural engineering|Electrical engineering|Remote sensing
Recommended Citation
Shropshire, Geoffrey John, "Weed detection in row crops using the red near-infrared reflectance ratio and frequency transforms of video images" (1989). ETD collection for University of Nebraska-Lincoln. AAI8925260.
https://digitalcommons.unl.edu/dissertations/AAI8925260