How to segment color image(Skin lesion) with Unet and transfert learning?

Question

Samia OUKIL 2020-5-11

回答： Prasanna 2024-10-28，8:40

Hello, I am a beginner in deep learning! So,I have a medical image database (Skin lesion) to segment with U-net and transfer learning! For that, I downloaded " u-net-release-2015-10-02.tar.gz"(https://lmb.informatik.uni-freiburg.de/people/ronneber/u-net/) but I don't know how to segment my database!

Please, if you can help and guide me to segment with Unet and how to use transfer learning?

First you have to groundtruth labeling your data.

Defind your data,

Answer 1

Prasanna 2024-10-28，8:40

Hi Samia,

To segment skin lesion images using Unet, you can follow the below steps:

Setup the MATLAB environment, and extract the U-Net files from the contents of 'u-net-release-2015-10-02.tar.gz'.
Organize your images and masks into two folders: one for the images and another for the masks. Then, Preprocess the data and split data for training and validation
Create a U-Net model using built-in functions such as the 'unet' function. To use transfer learning, you can modify the U-Net architecture to include a pre-trained network such as the 'resnet50' as the encoder. Train the model with appropriate training options to segment skin lesions.

For more information on U-Net and transfer learning to U-Net, refer the following resources:

Hope this helps!