Image Edge Detection using Fuzzy Inference

Fuzzy Logic Controller

Features

• FIS approach considers the imprecision and vagueness of image features, making it an optimal solution for edge detection.
• Traditional filter-based methods, such as the Sobel, Prewitt, and Roberts operators, rely on the same filter values to determine the presence of an edge.
• FIS controller is easily understandable.
• Fuzzy Logic algorithms can be coded using less data, so they do not occupy a huge memory space

Comparison

(a) Original Test Image; Edges detected using (b) Sobel’s filter (c) Robert’s filter (d) Prewitt’s filter (e) Canny Edge Detector (f) Fuzzy Logic Controller

Results

Result obtained by varying hyperparameters

(a) Original Test Image, Image edges detected for different values of defuzzification parameter,α (b) α=0.9 and (c) α =0.6

Application for MRI Medical Images of Brain

(a) MRI Image of Brain with Tumor (b) Edge detection of Tumor using Python (c) Edge detection of Tumor using MATLAB

Download images:

Download image from this site. https://cdn.pixabay.com/photo/2016/01/08/05/24/sunflower-1127174__340.jpg

Requirements

• Python Version >3.0

Libraries used:

• cv2
• numpy
• math
• pandas
• matplotlib.pyplot

Citation

If you found our model useful in your research, please consider starring ⭐ us on GitHub and citing 📚 us in your research!

@INPROCEEDINGS{10150762,
  author={Pandey, Aman Kumar and Chatla, H R S S Nagavalli and Pandya, Margi and Farhan M A, Aneesa and Rana, Ankur Singh},
  booktitle={2023 International Conference on Recent Advances in Electrical, Electronics & Digital Healthcare Technologies (REEDCON)}, 
  title={Image Edge Detection Using Fuzzy Logic Controller}, 
  year={2023},
  volume={},
  number={},
  pages={508-513},
  doi={10.1109/REEDCON57544.2023.10150762}}