Development of Inter-Leaves Weed and Plant Regions Identification Algorithm using Histogram of Oriented Gradient and K-Means Clustering

This paper proposes a weed detection mechanism, where the carrot leaves are segmented from the weeds (mostly Chamomile). In the early stage, both weeds and carrot leaves are intermixed with each other and have similar color texture. This makes it difficult to identify without the help of the domain experts. Therefore, it is essential to remove the weed regions so that the carrot plants can grow without any interruptions. The process of identifying the weeds become more challenging when both plant and weed regions overlap (inter-leaves). The proposed method takes account of this problem and breaks down the identification mechanism into three major components: Image Segmentation, Feature Extraction, and Classification. In the Image Segmentation stage, K-Means clustering is applied to select the images that will be used for the identification purpose. Next, in the Feature Extraction stage structural information of the weed and leaves will be extracted from the lower unit images. Furthermore, to extract the information from the Region of Interest (ROI), Histogram of Oriented Gradient (HoG) is used to locate and label all the weed and carrot leaves regions. In the Classification stage, Support Vector Machine (SVM) analyzes all the information and labels the regions. This method of weed detection is effective as it automates the identification process and fewer herbicides will be used, which in-turn benefits the environment. The proposed method successfully classifies the plant regions at a success rate of 92% using an open dataset and outperformed some of the previous approaches.