64-5 Plant Identification with Machine Vision.

See more from this Division: ASA Section: Climatology & Modeling
See more from this Session: Symposium--Use Of Remote Sensing For Crop and Pasture Statistics: I
Monday, November 4, 2013: 2:05 PM
Tampa Convention Center, Room 9
Share |

George E. Meyer, Biological Systems Engineering, University of Nebraska, Lincoln, NE
Undesirable plants negatively impact many ecosystems and cropping systems, causing significant losses both economically and environmentally.  Examples of such losses include altered stream flow in riparian areas, increased frequency of fires in rangelands and forests, fewer habitats with high species diversity and increased management for aesthetics and yields in natural areas and production systems, respectively.  An electronic plant species classification program has been developed and is currently being tested for identifying selected plant species found in Nebraska and fields of the Great Plains. The main identification process takes place through the presentation of a digital leaf image. The identification algorithm addresses detailed shape, edge serration, and a very detailed leaf venation network features. Specimens are imaged using a high-resolution, digital camera and special portable forward and back lighting systems. Three-dimensional canopy architecture and leaf features are also being evaluated using light field camera technology and laser scanning. Samples are supplied with plants grown in growth chambers, greenhouses, and spring/summer/fall field grown specimens. Species include various herbaceous plants that are native and non-native (e.g. velvet leaf, pigweed, downy brome, phragmites, common reed, leafy spurge, and cheat grass). A fuzzy-logic inference system (FIS) has been tested as an electronic taxonomy tool with good classification rates. The impact from the development of an electronic, real-time system to identify plant species would revolutionize the way plant populations and ecosystems are studied by researchers, and monitored by crop and land managers.

KEYWORDS:

Machine Vision, Plants, Species, Classification.

See more from this Division: ASA Section: Climatology & Modeling
See more from this Session: Symposium--Use Of Remote Sensing For Crop and Pasture Statistics: I