MUMBAI, India, July 13 -- Intellectual Property India has published a patent application (202621060167 A) filed by Bharati Vidyapeeth Deemed To Be University College Of Engineering, Pune on May 12, 2026, for Smart Hybrid Deep Learning System For Plant Leaf Disease Detection And Severity Classification.
Inventors include Ashwini Kanade; Dr. Priyanka Paygude; and Dr. Rajesh Prasad.
The application for the patent was published on July 10, 2026, under issue no. 28/2026.
Abstract: ABSTRACT SMART HYBRID DEEP LEARNING SYSTEM FOR PLANT LEAF DISEASE DETECTION AND SEVERITY CLASSIFICATION A smart hybrid deep learning system (100) and method (200) for plant leaf disease detection and severity classification are disclosed. The system comprises an image processing unit (102) configured to preprocess a plant leaf image using bilateral filtering to preserve lesion edge information while reducing environmental noise. A segmentation unit (104) comprising a parallel attention U-Net architecture generates a lesion mask corresponding to diseased regions through simultaneous channel attention and spatial attention processing. A feature extraction unit (106) extracts lesion-specific gradient-based texture features, structural binary pattern features, and color features. The extracted features are fused and processed using a hybrid classification unit (108) comprising a multi-residual convolutional neural network and a bidirectional gated recurrent unit network to identify plant disease classes. A severity estimation unit (110) determines disease severity levels including mild, moderate, and severe based on lesion characteristics. (To be published along with FIG. 4)
Disclaimer: Curated by HT Syndication.
