The main limitations of agricultural productivity are diseases and shortage of micro-nutrients especially in vegetables like bitter gourd. It is of essence that such stresses are detected early in an attempt to maximize the yield so that farming activities can find a place within the sustainability of the environment. The dataset that we have used in this paper is the image dataset of an olive leaf namely OLID-I subset of bitter gourd leaves that has a total of 785 high-resolution images divided into nine different classes, namely, healthy, disease, insect attack and nutrient deficiencies. Note, as the baseline, we use AgriCure-YOLOv8 model, trained on layered augmentations, with reported good performance of mAP score of 92.9% and a precision of 89.6% and a recall of 86.6%. Therefore, in our work we have QML algorithms in mind on leaf stress classification and using the result offered by the YOLOv8 baseline as the starting point we carry out scale-up analysis. Three models are compared and contrasted; the Quantum Convolutional Neural Network (QCNN) which is a Hierarchical Quantum Feature Extraction, the Variational Quantum Classifier (VQC) which is a direct quantum classification and a Hybrid Quantum-Classical Classifier (HQ-CC), in which the classical pre-classification is performed before the quantum classification. The obtained findings indicate that the quantum models are workable and can be deployed to handle small but complicated agronomic data and yield benefits in the realms of generalization performance and feature representations. In comparison to the YOLOv8 baseline accuracy, the paper presents quantum models, where the additional performance is demonstrated on identifying nutrient deficiencies, which is the most significant outcome that will be utilized in entering this new paradigm technology of smart agriculture, and in resource-poor environments, in particular, where plant stress is of paramount importance.In contrast to the current deep learning-based systems like the YOLOv8, the proposed quantum models have a higher generalization capability based on small and complex dataset and higher sensitivity to detect nutrient deficiencies. This means that QML is applicable in a resource-constrained agricultural setting.
The main limitations of agricultural productivity are diseases and shortage of micro-nutrients especially in vegetables like bitter gourd. It is of essence that such stresses are detected early in an attempt to maximize the yield so that farming activities can find a place within the sustainability of the environment. The dataset that we have used in this paper is the image dataset of an olive leaf namely OLID-I subset of bitter gourd leaves that has a total of 785 high-resolution images divided into nine different classes, namely, healthy, disease, insect attack and nutrient deficiencies. Note, as the baseline, we use AgriCure-YOLOv8 model, trained on layered augmentations, with reported good performance of mAP score of 92.9% and a precision of 89.6% and a recall of 86.6%. Therefore, in our work we have QML algorithms in mind on leaf stress classification and using the result offered by the YOLOv8 baseline as the starting point we carry out scale-up analysis. Three models are compared and contrasted; the Quantum Convolutional Neural Network (QCNN) which is a Hierarchical Quantum Feature Extraction, the Variational Quantum Classifier (VQC) which is a direct quantum classification and a Hybrid Quantum-Classical Classifier (HQ-CC), in which the classical pre-classification is performed before the quantum classification. The obtained findings indicate that the quantum models are workable and can be deployed to handle small but complicated agronomic data and yield benefits in the realms of generalization performance and feature representations. In comparison to the YOLOv8 baseline accuracy, the paper presents quantum models, where the additional performance is demonstrated on identifying nutrient deficiencies, which is the most significant outcome that will be utilized in entering this new paradigm technology of smart agriculture, and in resource-poor environments, in particular, where plant stress is of paramount importance.In contrast to the current deep learning-based systems like the YOLOv8, the proposed quantum models have a higher generalization capability based on small and complex dataset and higher sensitivity to detect nutrient deficiencies. This means that QML is applicable in a resource-constrained agricultural setting.
This study investigates the effectiveness of a hybrid quantum-classical framework for plant disease detection, with a focus on improving classification accuracy compared to conventional approaches. Image-based datasets of crop diseases were processed and encoded into variational quantum circuits (VQCs), which were then integrated with classical machine learning classifiers. The proposed model achieved an overall classification accuracy of 95.2%, outperforming a baseline convolutional neural network that recorded 91.4% accuracy on the same dataset. Performance evaluation was carried out using multiple metrics including precision, recall, and F1-score, with the hybrid approach showing consistent gains across all categories. Furthermore, the model demonstrated reduced overfitting and improved robustness when tested under noisy and high-dimensional input conditions. These results highlight the potential of quantum-assisted machine learning to significantly enhance crop disease detection, offering a scalable and accurate pathway for sustainable agriculture and precision farming.
Potato leaf diseases pose a serious challenge to agricultural sustainability and food security in regions such as West Bengal, India, where potato cultivation plays a vital socio-economic role. Although classical deep learning models demonstrate strong performance, they often suffer from overfitting, limited generalization, and heavy reliance on large labeled datasets, while manual inspection remains time-consuming and error-prone. This work proposes a hybrid Quantum Convolutional Neural Network (QCNN) integrated with 4-qubit parameterized quantum circuits (PQCs) for accurate classification of five major potato leaf disease categories, including nutrient deficiency, fungal, and viral infections. The model incorporates quantum activation layers, quantum softmax, and a focused loss function to enhance non-linear feature representation and multi-class separability. Grad-CAM is employed to improve the interpretability of predictions. Experiments conducted on a diverse, field-collected dataset from Bardhaman, Bankura, and Purulia districts achieve a classification accuracy of 99.94%, demonstrating superior generalization, robustness, and explainability compared to classical approaches. This study highlights the practical potential of quantum-enhanced AI for precision agriculture.
Plants and plant-produced vegetables are crucial for every age group due to their rich nutrient content, and tomatoes, as one of the leading vegetables, make a significant impact on human well-being.Tomatoes can easily be affected by environmental conditions and fungal infections, causing significant losses for farmers. Experimentation of Quantum machine learning and image processing creates stunning results in early disease detection in medical field. This research takes this guidance and endeavors to apply various quantum machine learning classification techniques with optimization to classify diseased tomato leaves. Quantum Neural Network (QNN), Quantum Convolution Neural Network (QCNN), and Quantum-Classical Hybrid Convolution Neural Network (QCHCNN) with Quantum Approximate Optimization Algorithm (QAOA), COBOLYA (Constrained Optimization by Linear Approximation and Simultaneous Perturbation Stochastic Approximation (SPSA) were experimented for tomato leaf disease detection to identify a timely and low-cost model. These classifiers achieved accuracies of 87%, 90%, and 99% respectively. The results suggest Quantum-Classical Hybrid Convolution Neural Network (QCHCNN) as the preferred choice for tomato leaf disease detection.
Detecting plant diseases is a crucial component of precision agriculture, necessary for maintaining economic viability, sustainable farming practices, food security, and preventing economic losses. Deep learning models such as Capsule Attention Networks, DenseNets, and Convolutional Neural Networks have demonstrated state-of-the-art performance in leaf-image-based disease classification. However, large-scale practical implementation in resource-constrained environments remains challenging due to overfitting, environmental insensitivity, extended training times, and high computational cost, along with the absence of a sustainable framework for real-world deployment. Emerging quantum neural networks (QNNs), integrated with quantum-influenced optimization algorithms, offer promising avenues to enhance practical applicability. Their capabilities for high-dimensional feature mapping, entanglement, and superposition improve generalization and convergence in hybrid models that combine quantum layers with classical architectures. Incorporating quantum modules into DenseNets, attention mechanisms, capsule networks, and other hybrid systems may further enhance accuracy, enable robust disease segmentation, and reduce training time. Nonetheless, challenges remain regarding real-time deployment on edge devices, efficient data encoding, limited explainability of quantum-driven features, and hardware constraints. This survey reviews classical, hybrid, and quantum-enhanced approaches, provides comparative insights, and identifies research gaps toward developing scalable, interpretable, and dependable frameworks for plant disease detection and diagnosis.
This paper presents a systematic empirical benchmarking of quantum machine learning (QML) applications within the under-explored domain of agricultural image classification. The authors implement and evaluate four hybrid quantum-classical architectures—including variational quantum classifiers (VQC), quanvolutional neural networks (QNN), and a ResNet-50 combined with a Quantum Support Vector Machine (QSVM)—against a purely classical EfficientNet baseline. The architectures are tested on three agricultural disease datasets: potato tuber disease, jujube insect pests, and paddy leaf rust. The experimental findings reveal that the ResNet-50+QSVM hybrid model achieves the optimal balance of classification accuracy and training stability, reaching a 93.4% mean accuracy. Crucially, the hybrid QSVM demonstrates extreme sample efficiency by requiring 41% fewer labeled training samples than the classical EfficientNet baseline to cross a 90% accuracy threshold, showing the capability of quantum feature maps in low-data regimes.
Chili (Capsicum annuum) is an economically important crop that is frequently affected by foliar diseases such as leaf curl, leaf spot, whitefly infestation, and yellowing, which cause substantial yield losses if not detected early. Classical convolutional neural networks (CNNs) have been widely applied to automated leaf-disease recognition, but they demand large labelled datasets, are computationally expensive, and their purely classical feature space limits discrimination between visually similar disease symptoms. This paper proposes a Hybrid Quantum-Classical Neural Network (HQ-CNN) that couples a lightweight classical convolutional feature extractor with a parameterised variational quantum circuit (VQC) acting as a trainable quantum feature map. RGB leaf images are pre-processed and passed through convolutional blocks to obtain a low-dimensional embedding, which is angle-encoded into a 4-8 qubit circuit; entangling variational layers followed by Pauli-Z measurement produce quantum features that are fed to a final softmax classifier. The complete network is trained end-to-end using the Adam optimiser for classical weights and the parameter-shift rule for quantum weights, and is evaluated on a five-class, publicly available chilli leaf disease dataset (Healthy, Leaf Curl, Leaf Spot, Whitefly, Yellowish). On this dataset, the proposed HQ-CNN attains 99.1test accuracy with only 24 quantum parameters and a total trainable-parameter count two to three orders of magnitude smaller than classical transfer-learning baselines (traditional CNN, VGG16, MobileNetV2, EfficientNetB0) reported in the literature, and an ablation study confirms that both the presence and the depth of the quantum layer contribute to this result. The results indicate that quantum feature fusion is a promising low-parameter-count direction for fine-grained agricultural image classification, particularly under limited-data regimes typical of plant pathology datasets.