|Evett, Steven - Steve|
|RUSH, CHARLES - TEXAS AGRILIFE RESEARCH|
Submitted to: Sensors
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/15/2014
Publication Date: 9/23/2014
Citation: Casanova, J.J., Oshaughnessy, S.A., Evett, S.R., Rush, C.M. 2014. Development of a wireless computer vision instrument to detect biotic stress in wheat. Sensors. 14:17753-17769.
Interpretive Summary: Plants can suffer stress from disease or lack of water. Diseased plants may not be able to use water efficiently or produce profitable yields. Therefore, withholding irrigation from diseased plants may save water. However, it is difficult to separate diseased and water stressed plants from healthy plants in a large size field. One method to aide in the detection of plant stress due to water or disease, is with a computer vision instrument. This device acquires digital images and separates information about the image into percent healthy and stressed plants. This information could be used by farmers to improve crop management. ARS scientists from Bushland analyzed digital images of diseased, healthy, water-stressed, and well-watered wheat throughout a growing season. Results showed that there was significant effect of crop stress on color derived from the images. The computer vision instrument was programmed onto a wireless microcomputer with a small digital camera and tested during a second growing season. This system could be used in the field for irrigation control.
Technical Abstract: Knowledge of soil water deficits, crop water stress, and biotic stress from disease or insect pressure is important for optimal irrigation scheduling and water management. While spectral reflectance and thermometry provide a means to quantify crop stress remotely, measurements can be cumbersome, expensive and affected by the amount of vegetative cover. Computer vision offers an inexpensive way to remotely detect crop stress independent of vegetation cover. This paper presents a technique using computer vision to determine water and disease stress in wheat. More specifically, digital images of differentially stressed wheat were segmented into soil and vegetation pixels using expectation maximization (EM). In the first season, the algorithm to segment vegetation from soil and distinguish between healthy and stressed wheat was developed and tested using digital images taken in the field and later processed on a desktop computer. In the second season, a wireless camera with near real-time computer vision capabilities, suitable for field use, was tested in conjunction with the conventional camera and desktop computer. For wheat, irrigated at different levels and inoculated with wheat streak mosaic virus at different times during the season, vegetation hue determined by the EM algorithm showed significant effects from irrigation level and infection status; vegetation cover showed significant effects from irrigation but not infection. Unstressed wheat had a lower hue (282.05) than the most stressed wheat (273.27) In the second season, the hue and cover measured by the wireless CVcomputer vision sensor showed significant effects from infection status (p=0.0014), as did the conventional camera (p less than 0.0001). This study shows that vegetation hue obtained through computer vision is a viable option for determining crop stress in irrigation scheduling.