RowDetr: End-to-End Row Detection Using Polynomials

2024-12-13

arXiv.org

AI Summary

In GPS-denied forest-understory farmland environments, crop row detection faces severe challenges including heavy occlusion, highly variable illumination, and irregular row geometry. To address these, we propose an end-to-end polynomial modeling framework: (1) we explicitly parameterize crop row geometry using low-order polynomials—a novel geometric representation; (2) we design PolyOptLoss, an energy-driven loss function that directly optimizes curve fitting in image space, significantly enhancing robustness to noisy annotations; and (3) we introduce a lightweight Transformer-based architecture and release FarmRow-6900, a diverse, large-scale farmland dataset comprising 6,900 annotated images. Our method achieves state-of-the-art accuracy, generalization, and real-time performance—outperforming Agronav and RowColAttention across all metrics, with 6× faster inference speed, enabling efficient edge deployment and establishing a new benchmark for understory crop row detection.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Learning to Search
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work

Abstract

Crop row detection is essential for enabling autonomous navigation in GPS-denied environments, such as under-canopy agricultural settings. Traditional methods often struggle with occlusions, variable lighting conditions, and the structural variability of crop rows. To address these challenges, RowDetr, a novel end-to-end neural network architecture, is introduced for robust and efficient row detection. A new dataset of approximately 6,900 images is curated, capturing a diverse range of real-world agricultural conditions, including occluded rows, uneven terrain, and varying crop densities. Unlike previous approaches, RowDetr leverages smooth polynomial functions to precisely delineate crop boundaries in the image space, ensuring a more structured and interpretable representation of row geometry. A key innovation of this approach is PolyOptLoss, a novel energy-based loss function designed to enhance learning robustness, even in the presence of noisy or imperfect labels. This loss function significantly improves model stability and generalization by optimizing polynomial curve fitting directly in image space. Extensive experiments demonstrate that RowDetr significantly outperforms existing frameworks, including Agronav and RowColAttention, across key performance metrics. Additionally, RowDetr achieves a sixfold speedup over Agronav, making it highly suitable for real-time deployment on resource-constrained edge devices. To facilitate better comparisons across future studies, lane detection metrics from autonomous driving research are adapted, providing a more standardized and meaningful evaluation framework for crop row detection. This work establishes a new benchmark in under-canopy

Problem

• Detects crop rows in GPS-denied agricultural environments.
• Addresses occlusions, lighting, and crop row variability.
• Enables real-time deployment on resource-constrained devices.

Innovation

• End-to-end neural network for row detection
• Polynomial functions for precise crop delineation
• PolyOptLoss enhances learning robustness and stability