Intellectual Property India Publishes Patent Application for Hybrid Quantum–convolutional Neural Networks For Morphological Soil Image Classification: Benchmarking Variational Quantum Circuits Against Classical Architectures

MUMBAI, India, July 30 -- Intellectual Property India has published a patent application (202641088076 A) filed by Koneru Lakshmaiah Education Foundation; Mittapally Anusha; and Dr. Nirmalajyothi Narisetty on July 19, 2026, for Hybrid Quantum–convolutional Neural Networks For Morphological Soil Image Classification: Benchmarking Variational Quantum Circuits Against Classical Architectures.

Inventors include Mittapally Anusha; and Dr. Nirmalajyothi Narisetty.

The application for the patent was published on July 24, 2026, under issue no. 30/2026.

Abstract: This study proposes a Hybrid Quantum Convolutional Neural Network (HQ-CNN) for classifying seven soil types using EfficientNet-B0 and a 4-qubit variational quantum circuit. The model was evaluated on 1,189 original and 5,097 augmented soil images. The classical CNN achieved 96.72% accuracy, while the HQ-CNN achieved 75.78% using only eight quantum parameters. Although less accurate, the HQ-CNN demonstrated the ability of quantum circuits to capture soil texture features under strong parameter constraints. The study provides a foundation for future quantum-based soil image classification.________________________________________

Disclaimer: Curated by HT Syndication.