Welcome

Normalization of stain often describes the transferring of color distribution from target image to source image and has been used to examine images in biomedicine. It is believed that traditional stain normalisation creates a pixel-by-pixel colour plotting approach that can't consistently perform style modifications between image databases since it's based on a single reference image. Though its complex network structure leads to low effectiveness in computation and artefacts in the style transformation, which has limited the real-world application in theory this deep learning techniques may effectively address the problem of style change. Here, A quick and reliable network called StainNet is used to absorb the colour mapping between the source and target picture, and distillation learning is utilised to simplify deep learning techniques. The dataset used is the images normalized using StainNET, Modified StainNet can learn the colour mapping association from a whole dataset and adjust the color value automatically. Here the modified architecture is analysed using Image augmentor package which is used to increase the dataset artificially based on the exsisting data. The AUC curve obtained is 91 percentage before augmentation and 93 percentage after augmentation. The results obtained from the Histopathology datasets demonstrate that StainNet, when updated, may perform on par with deep learning-based techniques.