自动对焦显微镜图像数据集
收藏资源简介:
本研究构建了一个大规模的自动对焦显微镜图像数据集,包含数百万张标记失焦图像,涵盖密集、稀疏和极其稀疏的场景。数据集由病理组织和细胞样本组成,包括来自肺部、肝脏、前列腺以及宫颈脱落细胞等样本,采用Papanicolaou染色和苏木精-伊红(H&E)染色。数据集通过自动显微镜成像平台收集,并分为训练集、验证集和测试集,用于训练和评估提出的SparseFocus方法,该方法能够有效处理不同内容稀疏度的自动对焦问题。
This study constructed a large-scale autofocus microscopy image dataset containing millions of labeled out-of-focus images covering dense, sparse, and extremely sparse scenarios. The dataset comprises pathological tissue and cell specimens, including samples sourced from lung, liver, prostate, and cervical exfoliated cells, which were stained with both Papanicolaou stain and hematoxylin-eosin (H&E) stain. Collected via an automated microscopy imaging platform, the dataset is partitioned into training, validation, and test subsets, and is utilized to train and evaluate the proposed SparseFocus method, which effectively addresses the autofocus problem across varying content sparsity levels.

- 1SparseFocus: Learning-based One-shot Autofocus for Microscopy with Sparse Content国防科技大学 · 2025年



