{ "created": "2023-12-26 16:10:48", "updated": "2026-05-06 06:33:45", "id": "ebb37465-45ac-4aaf-81f1-8f8612e79e4d", "version": 10, "ds_topic": null, "title_cn": "基于深度学习框架的中国31个主要城市细粒度城市绿地映射数据集", "title_en": "A fine-grained urban green space mapping dataset for 31 major cities in China based on deep learning framework", "ds_abstract": "

UGS-1m产品提供基于深度学习（DL）框架生成的中国31个主要城市的细粒度UGS地图。该生成器是一个专为UGS提取而设计的全卷积网络（UGSNet），它集成了注意力机制以提高对UGS的辨别力，并采用点撕裂策略进行边缘恢复。鉴别器是一个完全连接的网络，旨在处理图像之间的域偏移。数据集详细说明如下：

\n

1. UGS-1m.zip：中国 31 个主要城市的细粒度 UGS 地图产品；

\n

2.UGSet.zip：支持和促进 UGS 研究的大型基准数据集；

\n

3. GUB_Data.zip：各城市的全球城市边界数据；

\n

4. GE_Imagery_DataFrame.zip：\".shp \"格式的谷歌地球图像网格数据，提供每个城市的图像组成；

\n

5. 其他以城市名称命名的 Zip 文件：各城市的谷歌地球图像。", "ds_source": "

1.在谷歌地球下载了覆盖中国 31 个主要城市 GUB 区域的共 2179 幅 Google Earth 图像，每个数据框的经度为 7′30′′，纬度为 5′00′′，空间分辨率接近 1.1 米。

\n

2.通过高分二号（GF2）卫星从中国广东省的142个样本区域采集4544幅512×512大小、空间分辨率接近1米的图像，给大范围的城市绿地测绘提供广泛的样本数据库，并为深度学习算法之间的比较提供一个基准。

\n

3.利用 2018 年全球城市边界（GUBs；Li 等人，2020 年）数据对每幅样本图像的城市区域进行遮挡以滤除非城市区域的绿地。", "ds_process_way": "

获得 UGS-1m 的主要步骤可归纳如下：

\n

1. 首先，在 UGSet 上对 UGSNet 进行预训练，以便为生成器获得良好的起始训练点；

\n

2. 在 UGSet 上进行预训练后，判别器负责通过对抗训练使预训练的 UGSNet 适应不同的城市；

\n

3. 最后，使用 2179 幅谷歌地球图像获得中国 31 个主要城市的 UGS 结果（UGS-1m），数据帧的经度为 7'30\"，纬度为 5'00\"，空间分辨率接近 1. 1 米。

", "ds_quality": "

评价指标包括 OA、Pre、Rec 和 F1。可以看出，在五个验证城市中，所有城市的平均 OA 为 87.56%，而每个城市的 OA 都高于 85%。其中，长春的 OA 最高，为 90.62%，而北京的 OA 最低，也达到了 85.86%，说明不同城市的 UGS 结果基本良好。从 F1 分数来看，广州的 F1 分数最高，为 81.14%，其次是北京和长春，分别为 79.23% 和 77.10%。虽然武汉和拉萨的 F1 分数相对较低，分别为 67.71 % 和 59.85 %，但最终 UGS 结果的平均 F1 分数也达到了 74.86 %。此外，76.61 % 的平均 Rec 也表明 UGS 提取结果的漏检率相对较低，这在应用中非常重要。总之，经过在多个不同城市的定量验证，UGS-1m 的可用性得到了初步证明。", "ds_acq_start_time": null, "ds_acq_end_time": null, "ds_acq_place": "中国", "ds_acq_lon_east": 135.01666666666668, "ds_acq_lat_south": 3.8666666666666667, "ds_acq_lon_west": 73.66666666666667, "ds_acq_lat_north": 53.5, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "login-access", "ds_total_size": 426074410310, "ds_files_count": 36, "ds_format": "", "ds_space_res": "", "ds_time_res": "", "ds_coordinate": "无", "ds_projection": "", "ds_thumbnail": "ebb37465-45ac-4aaf-81f1-8f8612e79e4d.png", "ds_thumb_from": 2, "ds_ref_way": "", "paper_ref_way": "", "ds_ref_instruction": "", "ds_from_station": null, "organization_id": "0a4269e1-65f4-45f1-aeba-88ea3068eebf", "ds_serv_man": "敏玉芳", "ds_serv_phone": "0931-4967596", "ds_serv_mail": "ncdc@lzb.ac.cn", "doi_value": "", "subject_codes": [ "170.45" ], "quality_level": 3, "publish_time": "2023-12-27 11:00:13", "last_updated": "2026-01-14 10:56:02", "protected": false, "protected_to": null, "lang": "zh", "cstr": "https://cstr.cn/31253.11.sciencedb.07049", "i18n": { "en": { "title": "A fine-grained urban green space mapping dataset for 31 major cities in China based on deep learning framework", "ds_format": "", "ds_source": "

1. A total of 2179 Google Earth images covering the GUB area of 31 major cities in China were downloaded from Google Earth. Each data box has a longitude of 7 ′ 30 ′ 'and a latitude of 5 ′ 00 ′', with a spatial resolution of nearly 1.1 meters.

\n

2. Collect 4544 images with a size of 512 × 512 and a spatial resolution of nearly 1 meter from 142 sample areas in Guangdong Province, China through GF2 satellite, providing a wide sample database for large-scale urban green space mapping and a benchmark for comparing deep learning algorithms.

\n

3. Use the 2018 Global Urban Boundaries (GUBs; Li et al., 2020) data to occlude urban areas in each sample image to filter out green spaces in non urban areas.", "ds_quality": "

The evaluation indicators include OA, Pre, Rec, and F1. It can be seen that among the five validated cities, the average OA of all cities is 87.56%, and the OA of each city is higher than 85%. Among them, Changchun has the highest OA rate at 90.62%, while Beijing has the lowest OA rate at 85.86%, indicating that the UGS results in different cities are generally good. In terms of F1 scores, Guangzhou has the highest F1 score at 81.14%, followed by Beijing and Changchun at 79.23% and 77.10%, respectively. Although the F1 scores of Wuhan and Lhasa are relatively low, at 67.71% and 59.85% respectively, the average F1 score of the final UGS results also reached 74.86%. In addition, the average Rec of 76.61% also indicates that the missed detection rate of UGS extraction results is relatively low, which is very important in applications. In summary, the usability of UGS-1m has been preliminarily demonstrated through quantitative verification in multiple different cities.", "ds_ref_way": "", "ds_abstract": "

The UGS-1m product provides fine-grained UGS maps of 31 major cities in China generated based on a deep learning (DL) framework. This generator is a fully convolutional network (UGSNet) designed specifically for UGS extraction, which integrates attention mechanism to improve the discriminative power of UGS and adopts point tearing strategy for edge recovery. The discriminator is a fully connected network designed to handle domain shifts between images. The detailed description of the dataset is as follows:

\n

1. UGS-1m.zip: A fine-grained UGS map product for 31 major cities in China;

\n

2. UGSet.rip: A large benchmark dataset that supports and promotes UGS research;

\n

3. GUB_Datazip: Global city boundary data for each city;

\n

4. GE_imagery_dataFrame.rip: Google Earth image grid data in \". shp\" format, providing image composition for each city;

\n

5. Other Zip files named after city names: Google Earth images of each city.

", "ds_time_res": "", "ds_acq_place": "China", "ds_space_res": "", "ds_projection": "", "ds_process_way": "

The main steps to obtain UGS-1m can be summarized as follows:

\n

Firstly, pre train UGSNet on UGSet to obtain a good starting training point for the generator;

\n

After pre training on UGSet, the discriminator is responsible for adapting the pre trained UGSNet to different cities through adversarial training;

\n

Finally, using 2179 Google Earth images, UGS results (UGS-1m) were obtained for 31 major cities in China, with a longitude of 7'30 \"and a latitude of 5'00\" in the data frame, and a spatial resolution close to 1 1 meter.

", "ds_ref_instruction": "" } }, "submit_center_id": "ncdc", "data_level": 0, "license_type": "CC BY 4.0", "doi_reg_from": "reg_outside", "cstr_reg_from": "reg_outside", "doi_not_reg_reason": null, "cstr_not_reg_reason": null, "is_paper_in_submitting": false, "ds_topic_tags": [ "城市绿地", "深度学习", "地球科学" ], "ds_subject_tags": [ "地理学" ], "ds_class_tags": [], "ds_locus_tags": [ "中国" ], "ds_time_tags": [], "ds_contributors": [ { "true_name": "石茜", "email": "shixi5@mail.sysu.edu.cn", "work_for": "中山大学地理科学与规划学院", "country": "中国" }, { "true_name": "刘梦熙", "email": "liumx23@mail2.sysu.edu.cn", "work_for": "中山大学地理与规划学院", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "石茜", "email": "shixi5@mail.sysu.edu.cn", "work_for": "中山大学地理科学与规划学院", "country": "中国" }, { "true_name": "刘梦熙", "email": "liumx23@mail2.sysu.edu.cn", "work_for": "中山大学地理与规划学院", "country": "中国" } ], "ds_managers": [ { "true_name": "石茜", "email": "shixi5@mail.sysu.edu.cn", "work_for": "中山大学地理科学与规划学院", "country": "中国" }, { "true_name": "刘梦熙", "email": "liumx23@mail2.sysu.edu.cn", "work_for": "中山大学地理与规划学院", "country": "中国" } ], "category": "生态" }