martinjolif/football-player-detection

该数据集包含从足球比赛中捕获的图像，拍摄于球场中部的摄像机位置。每张图像都包含将物体分类为四个类别的标注：1. 球；2. 守门员；3. 球员；4. 裁判。对于图像中的每个物体，都提供了一个边界框，格式包括：`x_center`（边界框中心的X坐标）、`y_center`（边界框中心的Y坐标）、`width`（边界框的宽度）和`height`（边界框的高度）。该数据集适用于目标检测等任务，可用于检测足球比赛图像或视频中的球员。检测球员可能具有挑战性，因为他们经常靠得很近，可能会部分遮挡彼此，并且可能与裁判混淆。

This dataset consists of images captured from a football match in a stadium, taken from a camera positioned near the middle of the pitch. Each image contains annotations classifying objects into four categories: 1. Ball; 2. Goalkeeper; 3. Player; 4. Referee. For every object in an image, a bounding box is provided, represented in the format: `x_center` (X-coordinate of the bounding box center), `y_center` (Y-coordinate of the bounding box center), `width` (Width of the bounding box), and `height` (Height of the bounding box). This dataset is suitable for tasks such as object detection, and can be used to detect players in football game images or videos. Detecting players can be challenging, as they are often close together, can partially occlude each other, and can be confused with referees.