Textural Imaging and Discriminant Analysis for Distinguishingweeds for Spot Spraying

Advanced computer vision and statistical methods were employed for identifying living plants from soil/residue background for two species of grasses (Shattercane, Green Foxtail) and two broadleaf species (Velvetleaf, RedRoot Pigweed) weeds. The excess green index method was used as a contrast enhancement for specifically identifyingplant from soil regions. Excess green classified plant and soil regions correctly over the entire three-week observationperiod with high accuracies (99% plus). Plant and soil binary images were derived from excess green images andprovided edge boundaries. These boundaries were used with corresponding gray scale images to extract four classicaltextural features for plants and soil: angular second moment, inertia, entropy, and local homogeneity. These features werederived from the co-occurrence matrix. Stepwise and canonical discriminant analyses were used to test the classificationperformance of the texture and excess green features. Discrimination models of local homogeneity, inertia, and angularsecond moment were found to classify grass and broadleaf categories of plants, with classification accuracies of 93 and85%, respectively. Classification accuracies of individual species only ranged from 30 to 77%. Soil classificationaccuracies were also high for textural feature algorithms (97%). The time required to produce tokensets ranged from 15 to20 s on a UNIX computer system. Additional time required for the system to reach a plant/soil classification ranged from5 to 10 s. This translated into an overall system response time of 20 to 30 s, with the preprocessing step constituting themajor part of the system response time.