---
title: NCSTP_A_Benchmark_Dataset_for_Non_Cooperative_Space_Target_Perception_v4
canonical_url: "https://www.modelscope.cn/datasets/Wind006/NCSTP_A_Benchmark_Dataset_for_Non_Cooperative_Space_Target_Perception_v4"
md_url: "https://www.modelscope.cn/datasets/Wind006/NCSTP_A_Benchmark_Dataset_for_Non_Cooperative_Space_Target_Perception_v4.md"
repository: Wind006/NCSTP_A_Benchmark_Dataset_for_Non_Cooperative_Space_Target_Perception_v4
chinese_name: "NCSTP：非合作空间目标感知基准数据集"
last_updated: 2026-04-29
license: CC-BY-4.0
storage_size: "17 GB"
domain:
  - image
tasks:
  - image-classification
  - image-segmentation
  - object-detection
size_scale:
  - 10k-1m
downloads: 29
stars: 0
---

# NCSTP_A_Benchmark_Dataset_for_Non_Cooperative_Space_Target_Perception_v4

> NCSTP_A_Benchmark_Dataset_for_Non_Cooperative_Space_Target_Perception_v4 - Wind006 在 ModelScope 开源的数据集。This upload is an unmodified copy of the original dataset (version 4), hosted solely for convenient access. The uploader is not the author. Original…

Wind006/NCSTP_A_Benchmark_Dataset_for_Non_Cooperative_Space_Target_Perception_v4 是 ModelScope 魔搭社区上的image-classification、image-segmentation、object-detection数据集，涉及 image 领域，存储大小 17 GB，采用 CC-BY-4.0 许可。

- **Repository**: Wind006/NCSTP_A_Benchmark_Dataset_for_Non_Cooperative_Space_Target_Perception_v4
- **License**: CC-BY-4.0
- **Tasks**: image-classification, image-segmentation, object-detection
- **Domain**: image
- **Storage size**: 17 GB
- **Downloads**: 29
- **Stars**: 0
- **Last updated**: 2026-04-29

Source: https://www.modelscope.cn/datasets/Wind006/NCSTP_A_Benchmark_Dataset_for_Non_Cooperative_Space_Target_Perception_v4

---

# NCSTP Dataset Mirror

## Notice

This repository is an unmodified mirror of the original **NCSTP: A large-scale dataset for non-cooperative space target perception** dataset, hosted solely for convenient access.

The uploader is **not** the dataset author and does not claim ownership of the dataset, annotations, paper, or related materials. All dataset files are kept unchanged from the original release.

Original dataset source: https://doi.org/10.6084/m9.figshare.28606754

Original paper: https://doi.org/10.1038/s41597-025-06056-8

Original authors: Yuxuan Liu, Chunjiang Bian, Hongbin Nie, Shi Chen, and Ziqian Yang.

License: CC BY 4.0

If you use this dataset, please cite the original paper and the original dataset source rather than citing this mirror as the primary source.

## Dataset Description

NCSTP is a large-scale synthetic benchmark dataset for non-cooperative space target perception. It is designed for computer vision tasks related to space targets, including space target detection, fine-grained target recognition, and satellite component segmentation.

The dataset contains simulated spaceborne images generated in Blender under different target sizes, poses, illumination conditions, and space backgrounds. According to the original paper, NCSTP contains 200,000 images. All images can be used for space target detection and recognition, and 100,000 satellite images include component segmentation annotations.

## Supported Tasks

This dataset is suitable for the following image tasks:

- Object detection
- Image classification / fine-grained recognition
- Image segmentation / component segmentation

The corresponding ModelScope task metadata is:

```yaml
image:
  object-detection:
    size_scale:
      - 10k-1m
  image-segmentation:
    size_scale:
      - 10k-1m
  image-classification:
    size_scale:
      - 10k-1m