A Stain Normalization Method for the White Blood Cell Image

In this paper, a new system is designed to translate dyeing style. The whole system includes blood smears preparation, image acquisition and segmentation, finally, training a dye style transmission network. In the system, we designed a new network based on Deep Learning for the leukocyte dyeing style translation. The network can automatically extract the features of sample images during training and reduces the influence of lighting, dyeing, and taking a photo by different devices. In addition, we also build two datasets, Base dataset and Target dataset.