Imaging System Analyzes Crop Health

Monday, August 03 2009

Farmers could soon reduce water, energy, and labor costs by using imaging technology that analyzes crop health. Opto-Knowledge Systems, Inc. (OKSI) of Torrance, CA has developed modeling and simulation software to work with a hyperspectral imager to analyze and monitor the health of crops on large, multi-acre farms. Hyperspectral imaging produces computer-enhanced graphic depictions based on a targeted object’s spectral absorption, reflectivity, and thermal radiance.

Through a Small Business Innovation Research (SBIR) Phase II contract, the Missile Defense Agency (MDA) funded OKSI to enhance its existing modeling and simulation algorithms to discern live warheads from clutter and countermeasures. Such technology would be used with hyperspectral imaging sensors on spacebased missile defense platforms.

After the MDA contract concluded, OKSI began considering commercialization potential for the technology. If the software could be used with a hyperspectral imaging system to detect and analyze spectral signatures of objects in space, what about monitoring something closer to home? OKSI decided that the best use of the software and imaging system combo on land was in agriculture, due to the difficulty of monitoring the health of acres of individual plants.

With the technology and application idea in hand, OKSI met with some farmers who grow the nearly 750,000 acres of cotton in the San Joaquin Valley of California, as well as with experts from the U.S. Department of Agriculture, local county government, and the University of California. What company researchers found was a problem endemic to farmland in Southern California: the land is desert that relies heavily on irrigation canals from the northern part of the state. Because of the natural lack of water, farmers tend to irrigate — typically mixing nutrients in with the water — and apply chemicals and insecticides on fixed schedules.

But the typical means of irrigating does not really take into consideration the balance between the health of the crops and environmental concerns. Crops will die if not watered or protected against insects. But if too many nutrients, chemicals, and pesticides are applied, the environment may be adversely impacted.

So OKSI gathered a team of biologists, entomologists, and other experts in the area to size up the problem and determine how OKSI’s technology could directly benefit the cotton growers of the San Joaquin Valley. They collected data — including specific spectral signatures and stress caused by lack of water, lack of nutrients, salinity of the soil, or insect infestation — on cotton plants grown in San Joaquin Valley.

The scientists determined the spectrally significant features of cotton are generally found in the visible spectrum and nearinfrared wavelengths, rather than mid-wave infrared, as is the case with missiles. OKSI and its team were then able to incorporate the information into the software suite for use with hyperspectral imagers and sensors.