In-Field Mapping of Grape Yield and Quality with Illumination-Invariant Deep Learning

2025-10-06

AI Summary

To address domain shift induced by illumination variations in vineyards, this paper proposes an end-to-end IoT-enabled robotic system for non-invasive, real-time, spatially resolved field mapping of grape yield and quality (i.e., sugar and acid content). The core methodological contribution is the Light-Invariant Spectral Autoencoder (LISA), a domain-adversarial spectral autoencoder that extracts illumination-invariant features from uncalibrated hyperspectral data, thereby significantly improving cross-temporal generalization of predictive models. The system integrates a deep learning pipeline for cluster detection and weight estimation, a hyperspectral analysis module, and the adaptive LISA framework. Experimental results demonstrate a fruit-cluster detection recall of 82% and a weight prediction R² of 0.76. Leveraging LISA, sugar and acidity prediction accuracy improves by over 20% relative to baseline methods. The system ultimately generates high-resolution, georeferenced maps of yield and quality distributions across the vineyard.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Remote Sensing / Geospatial AI
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

This paper presents an end-to-end, IoT-enabled robotic system for the non-destructive, real-time, and spatially-resolved mapping of grape yield and quality (Brix, Acidity) in vineyards. The system features a comprehensive analytical pipeline that integrates two key modules: a high-performance model for grape bunch detection and weight estimation, and a novel deep learning framework for quality assessment from hyperspectral (HSI) data. A critical barrier to in-field HSI is the ``domain shift" caused by variable illumination. To overcome this, our quality assessment is powered by the Light-Invariant Spectral Autoencoder (LISA), a domain-adversarial framework that learns illumination-invariant features from uncalibrated data. We validated the system's robustness on a purpose-built HSI dataset spanning three distinct illumination domains: controlled artificial lighting (lab), and variable natural sunlight captured in the morning and afternoon. Results show the complete pipeline achieves a recall (0.82) for bunch detection and a $R^2$ (0.76) for weight prediction, while the LISA module improves quality prediction generalization by over 20% compared to the baselines. By combining these robust modules, the system successfully generates high-resolution, georeferenced data of both grape yield and quality, providing actionable, data-driven insights for precision viticulture.

Problem

• Real-time mapping of grape yield and quality in vineyards
• Overcoming illumination-induced domain shift in hyperspectral imaging
• Developing robust deep learning for non-destructive agricultural assessment

Innovation

• IoT-enabled robotic system for grape yield mapping
• Deep learning framework for quality assessment from HSI
• Light-Invariant Spectral Autoencoder for illumination robustness