English

Abstract

Agriculture plays a significant role in overall GDP of any country. So, there are several efforts those are being made to develop and increase the crop yield. In any of crop, weed i.e. (unwanted crops) is the major concern that may leads to poor production of crop. Therefore, an automatic crop monitoring system is required to monitor the weed. This system will help the farmers to monitor the crops, in gradual fashion, once it has been cultivated. Then specific Vegetation Index (VI) would be applied to locate all green portions in the image that would be part of early wheat crop or weed patches. We used Object Based Image Analysis (OBIA) algorithm to detect the weed patches in RGB and Multispectral imagery captured by UAV at 30-60 m altitude is used to acquire the images of wheat fields. Once the weed patches successfully identified from between the crop rows and within the crop rows then connected component-based classification technique is used that successfully classify the detected object either the object is related to weed patches or actual crop. The core objective of this work is to lessen the human involvement and to introduce the latest techniques and computation technology, peculiarly to identify the weed patches within the crop rows as well as between the crop rows in wheat field. Moreover, exploitation of UAV technology is also the core objective that will significantly provide the site specifically herbicides spraying