Carving and Classification of Crops Weed Using Area Threshold

In the field of agriculture, weed is an unwanted plant. It consumes the resources of crop such as water, pesticides, space and grows at the superfluous place. These unwanted plants should be destroyed with the reliable method at reasonable price. In traditional method, farmers use more amounts of similar pesticides all over the field and it results in decrease of the soil fertility. Different weeds needs particular pesticide, use of same pesticide to all the weeds is less efficient. In this paper we provide framework to classify crops and weed using image processing techniques. Image slicing, labelling, and calculating the areas of the labelled objects are the components of our framework. We demonstrate the results for different weed areas and crop areas. The classification is satisfactory when the area of weed is less than crop. Our framework can be used in real time spraying machine to control the growth of weeds. We simulate the results using MATLAB2012a.