- Project name
- Cell segmentation from bright field microscopic image by using Convolutional Neural Network.
- MSc. Ali Ghaznavi
- Description of the activities of the project
- Image object segmentation can be defined as a procedure to localize a region of interest (ROI) in an image by utilizing artificial inteligence methods. Detection and Segmentation Process can be involved in separating an image foreground from its background, by classification or clustering colour or shape similarities between pixel regions.
- The main goal of the project focus on cell boundary detection and segmentation in bright-field light microscopy images based on machine learning (ML) and deep learning (DL) methods. In order to accomplish this aim, we will apply CNN based methods with different architecture to find most suitable architecture for our dataset and segment image pixels into the background and foreground region.
- The students will work with bright-field light microscope to collect data from MG63 and HeLa living human cells and develop efficient algorithm for detection and segmentation phase.
- We expect that the proposed method should be invariant to cell size, orientation and illumination conditions and have efficient performance in poorly defined cell boundary and noisy images with inhomogeneous background which are very common problem in microscopic image processing system.
- Number of students - 2
- high school students students - NO
- university students - YES
- Prepositions (what should the students know): Candidates should be familiar working with Python programming language and well knowledge on Keras, Tensorflow and Pandas library. Candidate should be experienced working with machine learning and Deep Learning methods.
Academic and University Center
Zamek 136, Nove Hrady