Tango Spacecraft Wireframe Dataset Model for Line Segments Detection

<strong>Reference Paper:</strong> <strong>M. Bechini, M. Lavagna, P. Lunghi, Dataset generation and validation for spacecraft pose estimation via monocular images processing, Acta Astronautica 204 (2023) 358–369</strong> M. Bechini, P. Lunghi, M. Lavagna. "Spacecraft Pose Estimation via Monocular Image Processing: Dataset Generation and Validation". In 9th European Conference for Aeronautics and Aerospace Sciences (EUCASS) <strong>General Description:</strong> The "<em>Tango Spacecraft Wireframe Dataset Model for Line Segments Detection</em>" dataset here published should be used for line detection and segmentation tasks. It is split into 30002 train images and 3002 test images representing the Tango spacecraft from Prisma mission, being the only publicly available dataset of synthetic space-borne images tailored to line detection tasks (up to our knowledge). The label of each image gives the reprojection of a simplified wireframe model of Tango on the image plane split into lines. The labels are written following the Wireframe Model format. The "<em>Tango Spacecraft Wireframe Dataset Model for Line Segments Detection</em>" is also the largest dataset with wireframe annotations available up to date. More information on the dataset split and on the label format are reported below. <strong>Images Information:</strong> The dataset comprises 30002 synthetic grayscale images of Tango spacecraft from Prisma mission that serves as train set, while the test set is formed by 3002 synthetic grayscale images of Tango spacecraft from Prisma mission in PNG format. About 1/6 of the images both in the train and in the test set have a non-black background, obtained by rendering an Earth-like model in the raytracing process used to define the images reported. The images are noise-free to increase the flexibility of the dataset. The illumination direction of the spacecraft in the scene is uniformly distributed in the 3D space in agreement with the Sun position constraints. <br> <strong>Labels Information:</strong> Labels in the Wireframe dataset format are here provided in separated JSON files. The files are formatted per each image as in the following example: width : 98 # width in pixels (int) of the current image height : 176 # height in pixels (int) of the current image lines : [[line1], [line2], ..., [lineN]] # list of lines in each image filename : tango_img_866.png # string with image name and format Per each line (line1, ... , lineN) in lines, the format is [x0, y0, x1, y1]. (x0, y0) are the coordinates (float) of the line starting point in the image reference frame (x pointing right and y pointing down with origin located in the top-left corner of the image).<br> (X1, y1) are the coordinates (float) of the line ending point in the image reference frame (x pointing right and y pointing down with origin located in the top-left corner of the image). Note that the starting point is assumed to be the left-most endpoint (lower x coordinate in image reference frame) of each line. In the case of vertical lines, the starting point is the upper-most endpoint (lower y coordinate in image reference frame) of each line. <strong>VERSION CONTROL</strong> <strong>v1.0</strong>: All the images (both for train and test) have different resolutions, with Tango always centered in the image. The height of the images is in the range 19 - 352 pixels, while the width is in the range 16 - 336 pixels. The height over width ratio spans from 0.34 to 3.25. <strong>v2.0</strong>: This version contains all the images of v1.0 in the .zip folder named <em>Tango_WF.zip</em>, while in the .zip folder named <em>Tango_WF_fullscale.zip</em> there is the dataset (both train and test) of full scale images. These images have width=height=1024 pixels. The position of tango with respect to the camera is randomly selected from a uniform distribution, but it is ensured the full visibility in all the images. The labels for the wireframe are in the same format of v1.0. Note: the dataset in v1.0 is obtained by cropping the fullscale images in v2.0 and by properly rescaling the wireframe annotations. Note: this dataset contains the same images of the <em>"Tango Spacecraft Dataset for Region of Interest Estimation and Semantic Segmentation</em><em>"</em> v1.0 (DOI: https://doi.org/10.5281/zenodo.6507863) and also "<em>Tango Spacecraft Dataset for Monocular Pose Estimation</em>" v1.0 (DOI: https://doi.org/10.5281/zenodo.6499007) and they can be used together by combining the annotations of the relative pose and the ones of the reprojected wireframe model of Tango, with also the ones of the ROI. <strong>These three datasets give the most comprehensive dataset of space borne synthetic images ever published</strong> (up to our knowledge).<strong>参考文献：</strong> <strong>M. Bechini, M. Lavagna, P. Lunghi, 基于单目图像处理（monocular images processing）的航天器位姿估计数据集生成与验证, 《航天学报》(Acta Astronautica) 204 (2023) 358–369</strong> M. Bechini, P. Lunghi, M. Lavagna. 《基于单目图像处理的航天器位姿估计：数据集生成与验证》，发表于第9届欧洲航空与航天科学大会（European Conference for Aeronautics and Aerospace Sciences, EUCASS）。 <strong>总体描述：</strong> 本次发布的「用于线段检测的探戈航天器线框数据集模型（Tango Spacecraft Wireframe Dataset Model for Line Segments Detection）」可应用于线段检测与分割任务。该数据集分为30002张训练图像与3002张测试图像，均源自Prisma任务中的探戈航天器；据我们所知，这是目前公开可用的、专为线段检测任务打造的合成航天图像数据集。每张图像的标签对应探戈航天器简化线框模型（Wireframe Model）在图像平面上的重投影结果，并以线段形式进行划分。标签遵循线框模型格式进行编写。该数据集也是目前公开可用的、带有线框标注的最大规模数据集。有关数据集划分与标签格式的更多详细信息将在下文说明。 <strong>图像信息：</strong> 本数据集包含30002张合成灰度图像作为训练集，均为Prisma任务中的探戈航天器；测试集则由3002张PNG格式（PNG）的合成灰度探戈航天器图像组成。训练集与测试集中约1/6的图像拥有非黑色背景，背景是通过在生成图像的光线追踪（raytracing）过程中渲染类地模型得到的。所有图像均无噪声，以提升数据集的通用性。场景中航天器的光照方向在符合太阳位置约束的三维空间内均匀分布。 <strong>标签信息：</strong> 本数据集提供遵循线框数据集格式的标签，存储于独立的JSON文件（JSON）中。每个图像对应的标签文件格式示例如下： width : 98 # 当前图像的像素宽度（整数类型） height : 176 # 当前图像的像素高度（整数类型） lines : [[line1], [line2], ..., [lineN]] # 单张图像内的所有线段列表 filename : tango_img_866.png # 图像名称与格式的字符串 对于每条线段（line1, ..., lineN），其格式为`[x0, y0, x1, y1]`。其中`(x0, y0)`为线段起点在图像参考系中的坐标（浮点数类型），图像参考系以左上角为原点，x轴向右，y轴向下；`(x1, y1)`为线段终点在图像参考系中的坐标（浮点数类型）。注：默认将线段最左侧的端点（图像参考系中x坐标更小的端点）作为起点；若为垂直线段，则将最上方的端点（图像参考系中y坐标更小的端点）作为起点。 <strong>版本控制</strong> <strong>v1.0：</strong> 所有图像（训练集与测试集）分辨率各不相同，探戈航天器始终居中显示。图像高度范围为19至352像素，宽度范围为16至336像素，高宽比范围为0.34至3.25。 <strong>v2.0：</strong> 该版本将v1.0的所有图像收纳于名为`Tango_WF.zip`的压缩包中；而名为`Tango_WF_fullscale.zip`的压缩包则包含全尺寸版本的数据集（训练集与测试集），全尺寸图像的宽高均为1024像素。探戈航天器相对于相机的位置从均匀分布中随机选取，且确保所有图像中航天器完全可见。线框标签格式与v1.0保持一致。注：v1.0版本的数据集是通过裁剪v2.0的全尺寸图像并对线段标注进行适当缩放得到的。注：本数据集与v1.0版本的「用于感兴趣区域（Region of Interest, ROI）估计与语义分割的探戈航天器数据集（Tango Spacecraft Dataset for Region of Interest Estimation and Semantic Segmentation）」（DOI: https://doi.org/10.5281/zenodo.6507863）以及v1.0版本的「用于单目位姿估计的探戈航天器数据集（Tango Spacecraft Dataset for Monocular Pose Estimation）」（DOI: https://doi.org/10.5281/zenodo.6499007）包含相同的图像，可通过结合相对位姿标注、探戈航天器重投影线框模型标注以及感兴趣区域标注进行联合使用。<strong>据我们所知，这三个数据集共同构成了迄今为止发布的最全面的航天合成图像数据集。</strong>