CELL SEGMENTATION AND COUNTING USING U-NET ARCHITECTURE

Abstract

The purpose of this thesis is to create a deep learning application for solving the problem of detection the cells in some images. Artificial intelligence is becoming increasingly important in the field of biology, imagery, and medicine, as it can aid in processes that are difficult to do by humans. Image analysis tasks can be performed in a less prone to error way by introducing these algorithms, in such a way of avoiding issues with biological variance, variations in contrast or brightness, slide preparation, cell anomalies, arrangements, etc. In this work I am going to use U-Net to achieve cell segmentation, by training the neural network on our dataset of cell images.