A Novel Meta-analysis and Classification of Herbal Medicinal Plant Raw Materials for Food Consumption Prediction Using Hybrid Deep Learning Techniques Based on Augmented Reality in Computer Vision

The medicinal use of therapeutic plants in the diet is a long-standing historical practice across the world, particularly in India. Phytochemicals found in medicinal herbal plants have a significant role in food preparation. Because of their appearance, vital medicinal plants must be evaluated and examined before being consumed. Herbs may help prevent and control cardiac conditions, tumors, and mellitus. It may also help prevent blood clots and have anti-inflammatory and anti-tumor properties. The research article’s proposed new methodology has implemented a hybrid model that may be used to incorporate deep learning techniques such as ResNet, Inceptionv3, ViT, and Xception for the classification and meta-analysis of herbal medicinal plants for food consumption. The system’s information includes medicinal herbs that might be regarded as raw materials for food intake. The images of herbal plants in the collection were captured in a variety of settings with no regard for the surroundings. The hybrid model’s accuracy is 89%, 91%, 95%, and 98% by ResNet, Inceptionv3, ViT Architecture, and Xception, respectively, using mixed model privileges to achieve correct results by comparing several algorithms. Here, the system is meta-analyzed for optimization using the ViT architecture. The model’s outcome is shown utilizing augmented reality to improve user involvement. Augmented reality allows consumers to see the results in 3D and scan the medicinal plants using mobile devices, increasing their convenience. Augmented reality is implemented via computer vision, which establishes the environment for image processing and integrating deep learning algorithms. The suggested study improves usability and provides an enhanced way to investigate diverse therapeutic herbal plants that may be consumed as food.