Indonesian Journal of Electrical Engineering and Computer Science Vol. 30, No. 1, April 2023, pp. 356~365 ISSN: 2502-4752, DOI: 10.11591/ijeecs.v30.i1.pp356-365  356 Journal homepage: http://ijeecs.iaescore.com Hybrid deep-spatio textural feature model for medicinal plant disease classification Margesh Keskar, Dhananjay D. Maktedar Department of Computer Science and Engineering, Guru Nanak Dev Engineering College, Visvesvaraya Technological University Belagavi, Belagavi, India Article Info ABSTRACT Article history: Received Jul 13, 2022 Revised Nov 30, 2022 Accepted Dec 5, 2022 The high-pace rise in the demands of medicinal plants towards pharmaceutical significances as well as the different ayurvedic or herbal remedials have forced agro-industries However, rising plant disease cases have limited the cumulative growth and hence both volumetric production as well as quality of medicine. In this paper a first of its kind evolutionary computing driven ROI- specific hybrid deep-spatio temporal textural feature learning model is developed for medicinal plant disease detection (HDST-MPD). To alleviate any possible class-imbalance problem, HDST-MPD model at first applied firefly heuristic driven fuzzy C-means clustering to retrieve ROI-specific RGB regions. Subsequently, to exploit maximum possible deep spatiotemporal textural features, it applied gray-level co-occurrence matrix (GLCM) and AlexNet transferable network. Here, the use of multiple GLCM features helped in exploiting textural feature distribution, while AlexNet deep model yielded high-dimensional features. These deep-spatio temporal textural feature (deep-STTF) features were fused together to yield a composite vector, which was trained over random forest ensemble to perform two-class classification to classify each sample medicinal image as normal or diseased. Depth performance assessment confirmed that the proposed model yields accuracy of 98.97%, precision 99.42%, recall 98.89%, F-measure 99.15%, and equal error rate of 1.03%, signifying its robustness towards real-time medicinal plant disease detection and classification. Keywords: AlexNet Gray-level co-occurrence matrix Heuristic driven segmentation Hybrid deep-STTF feature learning Medicinal plant disease detection This is an open access article under the CC BY-SA license. Corresponding Author: Margesh Keskar Department of Computer Science and Engineering, Guru Nanak Dev Engineering College Visvesvaraya Technological University Belagavi Belagavi, Karnataka, India Email: mskeskar38@gmail.com 1. INTRODUCTION There are a large number of herbs including Karpooravalli (Coleus ambonicus), Podina (Mentha arvensis), Neem (Adidirachta indica), Thudhuvalai (Solanum trilobatum), and Basil (Ocimum sanctum). Naeem et al. [1], whose leaves are used for medicine manufacturing or herbal (or ayurvedic) remedial. Noticeably, these plants possess vital anti-bacterial efficiency so as to improve natural immunity [2], [3]. These key facts infer that the medicinal plants are of great significance to support pharmaceutical industry as whole while providing large remedial for ayurvedic and herbal treatments [1]-[4]. Its employees hararchical cluster analysis, fuzzy principle component analysis and linear discriminant analysis [5] like India use aforesaid medicinal plants directly for the different purposes like ayurvedic treatment or herbal remedials [6], [7]. In sync with above discussed significance of medicinal plants in human-life, numerous efforts have been made globally to increase plant’s productivity as well as yields efficiency [7]-[10]. In this reference, the