Office for Technology Commercialization
http://www.research.umn.edu/techcomm
612-624-0550

Automated Detection of Nitrogen Deficiency in Corn

Technology #20160058

Questions about this technology? Ask a Technology Manager

Download Printable PDF

Image Gallery
Nitrogen Deficiency in CornNitrogen Deficiency AlgorithmNitrogen Deficiency
Categories
Researchers
Nikolaos Papanikolopoulos, PhD
Professor, Computer Science & Engineering
External Link (www.cts.umn.edu)
David Mulla, PhD
Professor, Soil, Water, and Climate
External Link (www.swac.umn.edu)
Dimitrios Zermas, PhD candidate
Computer Science and Engineering
Daniel Kaiser, PhD
Associate Professor, Soil, Water, and Climate
External Link (www.swac.umn.edu)
Vassilios Morellas, PhD
Research Staff, Director, Computer Science and Engineering
External Link (www.cts.umn.edu)
Michael Bazakos
Researcher, Computer Science and Engineering
Managed By
Kevin Nickels
Technology Licensing Officer 612-625-7289
Patent Protection

Provisional Patent Application Filed
Publications
Automation solutions for the evaluation of plant health in corn fields
Intelligent Robots and Systems (IROS) - 2015 IEEE/RSJ International Conference, 28 Sept.-2 Oct. 2015

Detect Nitrogen Depletion from Corn Leaves

A new data collection and algorithm system automatically detects nitrogen stress on the leaves of corn plants. Small-scale unmanned aerial vehicles (UAVs) collect high resolution images in the visual (RGB) spectrum by hovering over stressed areas at a low altitude. Computer vision RGB image analysis algorithms eliminate the need of a control/reference plant by analyzing the color and shape of the stressed leaves to determine potential nitrogen depletion. The algorithm identifies potential segments of the images that exhibit nitrogen deficiency, even when unhealthy leaves are heavily occluded by other (healthy or stressed) leaves. The algorithm determines the density of leaves affected by nitrogen stress and the severity of stress, leading to an accurate fertilizer recommendation to correct the deficiency. This low-cost solution can help farmers efficiently increase crop production in several ways: The method provides state of the field information on a regular basis and offers early detection of exact deficiencies present, and it determines whether or not a crop needs nitrogen fertilizer and estimates the amount of fertilizer needed and where to apply it. Furthermore, it prevents unnecessary fertilizer from entering the environment and polluting water sources.

Determine Nitrogen Deficiency Severity

Current field surveillance for estimating nitrogen stress is done either manually, which is time and labor intensive, or with the assistance of satellite imaging, which provides infrequent and costly information. Precision agriculture technologies using reflectance spectroscopy for assessing plant or soil nutrient contents (specifically nitrogen) rely strictly on RGB color and are unable to distinguish nitrogen deficiencies from other deficiencies. The computer vision technology used in this new system is able to distinguish nitrogen deficiencies from other deficiencies and crop stresses. In addition, the algorithm can determine the type, extent and severity of nitrogen deficiencies based on the number and density of leaves affected by relying on color, texture and the derivative of changes across an image.

BENEFITS AND FEATURES:

  • Computer vision using high resolution RGB images
  • Detects nitrogen stressed leaves of plants
  • eference or control (non-stressed) leaves not needed
  • Small-scale UAV analysis of corn fields, done remotely and without intense labor
  • More accurate
  • Provides a spectral diagnosis based on density and number of leaves.
  • Determines whether or not crops need nitrogen fertilizer and estimates the amount needed and where to apply it
  • Prevents unnecessary fertilizer from entering the environment and polluting water sources

APPLICATIONS:

  • Corn crops
  • Soybean crops
  • Precision agriculture
  • Precision farming
  • Fertilizing

Phase of Development - Prototype