Skip to Main Content
 

Global Search Box

 
 
 

ETD Abstract Container

Abstract Header

Assessment of Corn Plant Population at Emergence from Processed Color Aerial Imagery

Wolters, Dustin Joseph
http://orcid.org/0000-0001-6857-4797
http://rave.ohiolink.edu/etdc/view?acc_num=osu1437666741

Abstract Details

2015, Master of Science, Ohio State University, Food, Agricultural and Biological Engineering.
As the application of remote sensing within the agricultural community continues to increase, the amount of data and management information acquired from layered sensing approaches does as well. This information source provides producers the ability to make important management decisions in a timely fashion. Crop emergence and health are at the forefront of many of these management decisions early in the growing season. This study investigates the development and application of an image processing and classification algorithm (image processing tool) to accurately detect, asses, and quantify corn plant population rates at emergence from high resolution and high quality remote sensed imagery. Validation and testing of the image processing tool using ground-truthed stand count data was completed for six test plots (from three aerial images) acquired from a production corn field, to demonstrate the accuracy and robustness of the tool for real-world applications. The algorithm provided an average R2 value of 0.44 and RMSE value of 47.66% per row, over the six test plots when compared against ground-truthed data for identification of emerging corn plants, using 3.3 cm spatial resolution aerial images. But, after taking into account the physically undetectable corn plants in each section, resulting from the early growth stage and low image resolution, the average R2 value improved to 0.90 with an average RMSE of 21.18%. The algorithm also improved to an accuracy of 96% when a much higher quality aerial image with higher spatial resolution, 0.53 mm, was analyzed, indicating even better algorithm performance when the proper quality and spatial resolution to distinguish corn plants within an image is available. The classification model also created the first ever quantified spatial visualization of plant population, extracted from remote sensed imagery, via an emergence map. The emergence map covered the entire area of the subject image and allowed for a spatial representation of corn population estimates at emergence, up through the intra-row scale, to support early season crop management decisions. Model validation results and emergence map creation provided supporting evidence of the algorithm’s robustness and ability to detect, asses, and quantify corn plant population rates at emergence extracted from high quality and high resolution remote sensed imagery.
Dr. Scott Shearer (Advisor)
213 p.
Agricultural Engineering
aerial photography; algorithms; image processing; remote sensing; corn; agriculture

Recommended Citations

Citations

  • Wolters, D. J. (2015). Assessment of Corn Plant Population at Emergence from Processed Color Aerial Imagery [Master's thesis, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1437666741

    APA Style (7th edition)

  • Wolters, Dustin. Assessment of Corn Plant Population at Emergence from Processed Color Aerial Imagery. 2015. Ohio State University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1437666741.

    MLA Style (8th edition)

  • Wolters, Dustin. "Assessment of Corn Plant Population at Emergence from Processed Color Aerial Imagery." Master's thesis, Ohio State University, 2015. http://rave.ohiolink.edu/etdc/view?acc_num=osu1437666741

    Chicago Manual of Style (17th edition)

osu1437666741
145
© 2015, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.