CoNSeP

该数据集名为ConSeP，包含了从结肠病患者中采集的41幅图像。这些图像按照3:1:1的比例被划分为训练集、验证集和测试集。每幅图像被裁剪成256×256大小的块，块与块之间有128像素的重叠。整个数据集的规模为41幅图像，每幅图像的分辨率为1000×1000。该数据集的任务是对细胞核进行分割。

The dataset named ConSeP consists of 41 images collected from colon disease patients. These images are partitioned into training, validation and test sets at a ratio of 3:1:1. Each image is cropped into 256×256 patches with a 128-pixel overlap between adjacent patches. The total scale of the dataset is 41 images, each with a resolution of 1000×1000, and the task of this dataset is nuclear segmentation.