jensky/Jensky

该数据集名为baseball-detection-2，专注于与棒球检测相关的计算机视觉任务。数据集包含4153张图像，棒球以YOLOv9格式进行标注。每张图像都经过以下预处理：自动方向调整（去除EXIF方向信息）、调整大小为640x640（拉伸）、通过自适应均衡化进行自动对比度调整。此外，为每张源图像创建了两个版本，应用了以下增强技术：随机裁剪图像的0%至15%、随机亮度调整（-25%至+25%）、随机曝光调整（-9%至+9%）、对5%的像素应用椒盐噪声。

This dataset, named baseball-detection-2, is focused on computer vision tasks related to baseball detection. It contains 4153 images with baseballs annotated in YOLOv9 format. Each image underwent the following pre-processing: auto-orientation of pixel data (with EXIF-orientation stripping), resize to 640x640 (Stretch), and auto-contrast via adaptive equalization. Additionally, two versions of each source image were created with the following augmentations: randomly crop between 0 and 15 percent of the image, random brightness adjustment of between -25 and +25 percent, random exposure adjustment of between -9 and +9 percent, and salt and pepper noise applied to 5 percent of pixels.