A Review on Weed Detection Using Image Processing

摘要

Objectives: To detect and map the specific weed to control the cost of farming, and enhance productivity. Methods analysis: There are various methods used for weed detection. Image processing methods are used along with artificial intelligence methods like ANN (Artificial neural networks) and Support Vector Machine (SVM) to identify the weed. Findings/application: Weeds are identified from the image, making use of image processing technique by shape, color and size feature using Support Vector Machine or Artificial Neural Network. Since hand labor is expensive and automatic weed controlling would be more feasible and the robotic weed controller can decrease or eliminate the need for chemicals.Keywords: Weed Detection, Image Processing, Artificial Neural Network, Support Vector Machine