• Jyoti Rathod
• Surbhi Maheshwari
• Dr. Subhash Chandra

Medicinal plants play a significant role in traditional healthcare systems, pharmaceutical industries, and sustainable agriculture. Accurate identification of herbal plants and early detection of plant diseases are essential for maintaining crop productivity, medicinal quality, and economic sustainability. Traditional manual methods of plant identification and disease diagnosis are often time-consuming, subjective, and dependent on expert knowledge, making them unsuitable for large-scale agricultural applications. Recent advancements in artificial intelligence, machine learning, and deep learning have provided effective alternatives for automated plant analysis. This research proposes a hybrid machine learning framework for herbal plant classification and leaf disease detection by integrating Gray-Level Co-occurrence Matrix (GLCM) texture features with deep features extracted from the MobileNetV2 convolutional neural network. The combined feature representation is classified using a Random Forest algorithm to improve accuracy and robustness. The proposed framework utilizes two independent image datasets consisting of medicinal herb species and various plant leaf diseases. Image preprocessing techniques, including resizing and normalization, are applied before feature extraction. GLCM captures texture characteristics such as contrast, homogeneity, energy, correlation, and dissimilarity, while MobileNetV2 extracts high-level semantic features representing leaf structure and visual patterns. Experimental evaluation demonstrates that the hybrid approach effectively distinguishes herbal species and diseased leaves with high classification performance. Furthermore, the system incorporates a rule-based recommendation module that provides fertilizer and treatment suggestions based on the identified disease category, enhancing its practical value for smart agriculture. The integration of handcrafted texture features and deep learning representations improves classification reliability and overcomes limitations associated with individual methods. The proposed intelligent framework contributes to precision agriculture by supporting farmers, herbal practitioners, and agricultural experts with accurate, scalable, and automated decision-making tools for sustainable medicinal plant management and crop protection.

Herbal plant classification, Plant disease detection, Machine learning, Deep learning, GLCM, MobileNetV2, Random Forest, Precision agriculture, Computer vision, Smart farming.

"A Hybrid Deep Learning Framework for Medicinal Herbal Plant Classification and Plant Disease Detection Using GLCM, MobileNetV2, and Random Forest", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 10, page no.a966-a975, October-2026, Available :http://www.jetir.org/papers/JETIR2610100.pdf

"A Hybrid Deep Learning Framework for Medicinal Herbal Plant Classification and Plant Disease Detection Using GLCM, MobileNetV2, and Random Forest", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 10, page no. ppa966-a975, October-2026, Available at : http://www.jetir.org/papers/JETIR2610100.pdf