Abstract:

Weed identification across the vegetable plantation is most challenging in the crop prediction based techniques and many researchers are understanding the study of identification of leaf and weed across the agriculture fields using many techniques through which the identifications of weeds in the vegetable plantation due to random plant spacing across the agriculture fields. Many traditional methods are implemented in order identify the week which are focused identifying weed directly in which for these manual techniques identifying weed across the vegetable leaves become more challenging aspects. This research paper provides a solution for the existing problem in integrating deep learning techniques in order to identify the weed plants across the vegetable plantation using CNN and advanced deep learning techniques like feature selection algorithms such as GABOR filter. Initially a trained Model was used over the data sets in order to draw the overlay by boundary boxes across the vegetable and weed leaves. The remaining space which was falling out of the overlay boundary boxes will be considered as weed through advanced detection techniques. In this way, the model focuses on identifying only the vegetables and thus avoid handling various weed species. Furthermore, this strategy can largely reduce the size of training image dataset as well as the complexity of weed detection, thereby enhancing the weed identification performance and accuracy. To extract weeds from the background, a colour index-based segmentation was performed utilizing image processing. The automation used for implementing identification of weeds is Deep learning (DL) and image processing (IP). Firstly, the Convolutional Neural Network (CNN) algorithm is used to recognize the weeds by drawing the bounding boxes around the green plants and the left-over parts are identified as crops. Later on, GABOR filter is used on same dataset, confusion matrix and accuracy are generated. Agri_data is the dataset used for training and testing data. By using the algorithms, we can identify whether they are weeds or crops. Accuracy of CNN and GABOR filter are compared for weed Identification and prediction.

Keywords: