用近红外影像提升停车区分割精度，构建中美可用数据集
A Pipeline and NIR-Enhanced Dataset for Parking Lot Segmentation
- 引入近红外通道+后处理技术，提升卫星图停车区识别
- 4通道输入结合后处理，达84.9% mIoU和96.3%像素准确率
- 适合城市规划、遥感分析人员，可复用于其他地物分割
停车位地图常需人工绘制，耗时且开源数据稀缺，尤其针对美国城市。本文提出利用近红外（NIR）波段增强卫星图像的停车区分割能力，并结合多种后处理技术。构建了两个各含12,617张图像-掩码对的数据集：一个为3通道（RGB），另一个为4通道（RGB + NIR）。在这些数据上训练了五种模型（OneFormer、Mask2Former、SegFormer、DeepLabV3、FCN）进行语义分割，区分停车与非停车区域。结果显示，尽管NIR通道需从低分辨率上采样，但其能有效提升精度，因停车场常被草地包围。进一步通过消除错误孔洞、简化边缘、去除道路与建筑轮廓等后处理步骤，显著提高性能。最佳模型为在4通道输入上训练的OneFormer，搭配后处理，达到84.9%的平均交并比（mIoU）和96.3%的像素级准确率。
原文摘要 · Abstract (English)
Discussions of minimum parking requirement policies often include maps of parking lots, which are time consuming to construct manually. Open source datasets for such parking lots are scarce, particularly for US cities. This paper introduces the idea of using Near-Infrared (NIR) channels as input and several post-processing techniques to improve the prediction of off-street surface parking lots using satellite imagery. We constructed two datasets with 12,617 image-mask pairs each: one with 3-channel (RGB) and another with 4-channel (RGB + NIR). The datasets were used to train five deep learning models (OneFormer, Mask2Former, SegFormer, DeepLabV3, and FCN) for semantic segmentation, classifying images to differentiate between parking and non-parking pixels. Our results demonstrate that the NIR channel improved accuracy because parking lots are often surrounded by grass, even though the NIR channel needed to be upsampled from a lower resolution. Post-processing including eliminating erroneous holes, simplifying edges, and removing road and building footprints further improved the accuracy. Best model, OneFormer trained on 4-channel input and paired with post-processing techniques achieves a mean Intersection over Union (mIoU) of 84.9 percent and a pixel-wise accuracy of 96.3 percent.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。