Rumex Weed Classification Using Region-Convolution Neural Networks Based-Colour Space Information

Weed detection is considered the gold standard in smart agriculture field. An automated detection of weed procedure is a complicated task, specifically detection of Rumex weed due to different real-world environ-mental conditions, including illumination, occlusion, overlapped, growth stage, and colours. Few works have done to classify Rumex weed using machine learning. However, the performance is still not at the level required for agriculture communities and challenges have not been solved. This work proposes Region-Convolutional Neural Networks (R-CNNs) and VGG16 model based on colour space information to classify Rumex weed from grass-land. The first step of the proposed method is to convert the images from RGB colour to HSV space colour due to robustness to lighting changes and removing shadows. Then, the images is divided into regions and we used selective search to select regions based on colour information. After that, the selective regions is entered into CNNs based imagenet weight and VGG16 model to extract features and the SVM classifier is used to classify these features into two classes (Rumex or Grass). This paper is investigated the effectiveness of our proposed method over real-world images under different conditions. The findings have shown that the proposed method superior comparing with other AI existing techniques. The results demonstrate that the proposed method has an excellent adaptability over real-world images in grassland environment. The findings provide valuable insight of using artificial intelligence in the agriculture field.