TowerDataset: A Heterogeneous Benchmark for Transmission Corridor Segmentation with a Global-Local Fusion Framework

2026-04-18

AI Summary

This study addresses key challenges in semantic segmentation of power transmission corridor point clouds, including scarcity of real-world data, difficulty in modeling long-range structural dependencies, and fine-grained discrimination of safety-critical components. To this end, the authors introduce TowerDataset, a large-scale heterogeneous benchmark comprising 661 scenes and 2.466 billion points, featuring the first complete-corridor annotations with 22 fine-grained semantic classes. They propose a global–local fusion framework that integrates whole-scene NoCrop training, prototype-based contrastive learning, patch-wise local processing, and geometric verification to effectively combine multi-scale features from Transformers or point cloud backbones. Experiments demonstrate significant improvements in recognizing rare and easily confused components on TowerDataset and two public datasets, validating both the benchmark’s challenge and the method’s robustness.

Technology Category

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

Application Category

• Security and Privacy: Data transparency and provenance
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Fine-grained semantic segmentation of transmission-corridor point clouds is fundamental for intelligent power-line inspection. However, current progress is limited by realistic data scarcity and the difficulty of modeling global corridor structure and local geometric details in long, heterogeneous scenes. Existing public datasets usually provide only a few coarse categories or short cropped scenes which overlook long-range structural dependencies, severe long-tail distributions, and subtle distinctions among safety-critical components. As a result, current methods are difficult to evaluate under realistic inspection settings, and their ability to preserve and integrate complementary global and local cues remains unclear. To address the above challenges, we introduce TowerDataset, a heterogeneous benchmark for transmission-corridor segmentation. TowerDataset contains 661 real-world scenes and about 2.466 billion points. It preserves long corridor extents, defines a fine-grained 22-class taxonomy, and provides standardized splits and evaluation protocols. In addition, we present a global-local fusion framework which preserves and fuses whole-scene and local-detail information. A whole-scene branch with NoCrop training and prototypical contrastive learning captures long-range topology and contextual dependencies. A block-wise local branch retains fine geometric structures. Both predictions are then fused and refined by geometric validation. This design allows the model to exploit both global relationships and local shape details when recognizing rare and confusing components. Experiments on TowerDataset and two public benchmarks demonstrate the challenge of the proposed benchmark and the robustness of our framework in real, complex, and heterogeneous transmission-corridor scenes. The dataset will be released soon at https://huggingface.co/datasets/tccx18/Towerdataset/tree/main.

Problem

• transmission corridor segmentation
• point cloud
• global-local fusion
• long-tail distribution
• fine-grained semantic segmentation

Innovation

• Transmission Corridor Segmentation
• Global-Local Fusion
• Point Cloud Semantic Segmentation
• Heterogeneous Benchmark
• Prototypical Contrastive Learning