MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641095251 A) filed by Adhiparasakthi Engineering College, Melmaruvathur on August 06, 2026, for Automated Quality Assessment For Vegetables Using Hybrid Deep Learning.
Inventors include Dr. C. Dhaya; Ms. V. Thamarai Selvi; A. Vanitha; Dr. R. Srivel; Mr. K. Chairmadurai; and Mr. G. Srinivasan.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: Post-harvest vegetable quality management is a critical challenge in global agricultural supply chains, with 45–60% of vegetables lost between farm and consumer due to inadequate and inconsistent quality grading. Traditional manual inspection and rule-based machine vision systems are fundamentally limited by inter-inspector variability, poor scalability, lack of interpretability, and heavy dependence on large annotated datasets. This invention proposes an Automated Quality Assessment System for Vegetables using a Hybrid Deep Learning Pipeline integrating DINOv2 self-supervised feature extraction, Convolutional Autoencoder (CAE) representation learning, K-Means and Deep Convolutional Embedded Clustering (DCEC) for unsupervised quality grouping, and fine-tuned transformer classifiers — Vision Transformer (ViT-B/16), Data-efficient Image Transformer (DeiT-B), and Convolutional Vision Transformer (CvT-21) — for three-class quality classification into Fresh, Moderate, and Spoiled grades. The system applies CLAHE-e
Disclaimer: Curated by HT Syndication.
