This paper presents a novel approach to agricultural disease diagnostics through the integration of Deep Learning (DL) techniques with Visual Question Answering (VQA) systems, specifically targeting the detection of wheat rust. Wheat rust is a pervasive and destructive disease that significantly impacts wheat production worldwide. Traditional diagnostic methods often require expert knowledge and time-consuming processes, making rapid and accurate detection challenging. We drafted a new, WheatRustDL2024 dataset (7998 images of healthy and infected leaves) specifically designed for VQA in the context of wheat rust detection and utilized it to retrieve the initial weights on the federated learning server. This dataset comprises high-resolution images of wheat plants, annotated with detailed questions and answers pertaining to the presence, type, and severity of rust infections. Our dataset also contains images collected from various sources and successfully highlights a wide range of conditions (different lighting, obstructions in the image, etc.) in which a wheat image may be taken, therefore making a generalized universally applicable model. The trained model was federated using Flower. Following extensive analysis, the chosen central model was ResNet. Our fine-tuned ResNet achieved an accuracy of 97.69% on the existing data. We also implemented the BLIP (Bootstrapping Language-Image Pre-training) methods that enable the model to understand complex visual and textual inputs, thereby improving the accuracy and relevance of the generated answers. The dual attention mechanism, combined with BLIP techniques, allows the model to simultaneously focus on relevant image regions and pertinent parts of the questions. We also created a custom dataset (WheatRustVQA) with our augmented dataset containing 1800 augmented images and their associated question-answer pairs. The model fetches an answer with an average BLEU score of 0.6235 on our testing partition of the dataset. This federated model is lightweight and can be seamlessly integrated into mobile phones, drones, etc. without any hardware requirement. Our results indicate that integrating deep learning with VQA for agricultural disease diagnostics not only accelerates the detection process but also reduces dependency on human experts, making it a valuable tool for farmers and agricultural professionals. This approach holds promise for broader applications in plant pathology and precision agriculture and can consequently address food security issues.

Municipal solid waste compost, the circular economy's closed-loop product often contains excessive amounts of toxic heavy metals, leading to market rejection and disposal as waste material. To address this issue, the study develops a novel approach based on: (i) utilizing plant-based biodegradable chelating agent, l-glutamic acid, N,N-diacetic acid (GLDA) to remediate heavy metals from contaminated MSW compost, (ii) comparative assessment of GLDA removal efficiency at optimal conditions with conventional nonbiodegradable chelator EDTA, and (iii) enhanced pre- and post-leaching to evaluate the mobility, toxicity, and bioavailability of heavy metals. The impact of treatment variables, such as GLDA concentration, pH, and retention time, on the removal of heavy metals was investigated. The process was optimized using response surface methodology to achieve the highest removal effectiveness. The findings indicated that under optimal conditions (GLDA concentration of 150 mM, pH of 2.9, retention time for 120 min), the maximum removal efficiencies were as follows: Cd-90.32%, Cu-81.96%, Pb-91.62%, and Zn-80.34%. This process followed a pseudo-second-order kinetic equation. Following GLDA-assisted leaching, the geochemical fractions were studied and the distribution highlighted Cd, Cu, and Pb's potential remobilization in exchangeable fractions, while Zn displayed integration with the compost matrix. GLDA-assisted leaching and subsequent fractions illustrated transformation and stability. Therefore, this process could be a sustainable alternative for industrial applications (agricultural fertilizers and bioenergy) and social benefits (waste reduction, urban landscaping, and carbon sequestration) as it has controlled environmental footprints. Hence, the proposed remediation strategy, chemically assisted leaching, could be a practical option for extracting heavy metals from MSW compost, thereby boosting circular economy.

Modeling the spatial variability and uncertainty of soil fertility parameters is crucial for sustainable agriculture but remains a challenge due to complex interactions between soil properties. Traditional models often assess individual parameters, such as pH or nitrogen (N), without considering their combined influence and uncertainty. This study develops a fuzzy logic and geoinformatics-based approach to simultaneously assess multiple soil fertility parameters. The model integrates 80 fuzzy rules to evaluate macro- and micronutrients, incorporating 250 soil samples analyzed using the PUSA Soil Test and Fertilizer Recommendation Meter (STFR). Experimental results showed soil fertility parameter ranges: pH (7.46–8.26), ECe (0.267–0.807 dS m⁻¹), organic carbon (0.24–0.56%), N (85.56–146.32 kg ha⁻¹), P (21.99–34.28 kg ha⁻¹), K (116.41–156.16 kg ha⁻¹), S (5.60–20.86 mg kg⁻¹), Fe (1.065–5.095 mg kg⁻¹), Mn (2.058–2.637 mg kg⁻¹), Zn (0.748–1.105 mg kg⁻¹), B (0.372–0.530 mg kg⁻¹), and Cu (0.230–0.788 mg kg⁻¹). The fuzzy model-derived fertility scores ranged from 41.55 to 52.60, with pH, organic carbon, nitrogen, phosphorus, potassium, and iron as critical parameters influencing fertility. Geostatistical kriging interpolation estimated fertility values at unsampled locations, generating a continuous, high-resolution soil fertility map for precision agriculture. Validation with crop yield data ranked suitability as: Pearl millet (0.919) > Mustard (0.890) > Wheat (0.863) > Barley (0.861). Multi-criteria decision analysis confirmed pearl millet as the most suitable crop based on fertility and yield potential. The study categorizes soil into low and moderate fertility zones across Jhunjhunu, Rajasthan, ensuring a systematic assessment for optimal nutrient management. By integrating fuzzy logic with GIS-based spatial modeling, this study enhances soil fertility classification, site-specific nutrient recommendations, and sustainable crop planning, reinforcing the role of fuzzy-GIS frameworks in precision agriculture.