Related Experiment Video
Updated: Apr 8, 2026
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
1.2K
Fuzzy Index to Evaluate Edge Detection in Digital Images
Felicitas Perez-Ornelas1, Olivia Mendoza1, Patricia Melin2
1School of Engineering, Autonomous University of Baja California, Tijuana, Baja California, Mexico.
Plos One
|June 27, 2015
Summary
This study introduces a novel fuzzy index (FI) for evaluating detected edges in digital images, avoiding information loss from binarization. The fuzzy index offers a more comprehensive edge evaluation compared to traditional metrics.
Area of Science:
- Computer Vision
- Image Processing
- Artificial Intelligence
Background:
- Existing edge evaluation metrics like Pratt's figure of merit (FOM), Jaccard's index (JI), and Dice's coefficient (DC) require image binarization.
- The binarization process inherent in current metrics leads to a loss of valuable image information, potentially compromising evaluation accuracy.
Purpose of the Study:
- To propose a new fuzzy index (FI) for evaluating detected edges in digital images.
- To overcome the limitations of existing metrics by eliminating the need for a binarization step.
Main Methods:
- Images are represented in their fuzzy form to process all detected edges.
- Calculations are performed using fuzzy set operators and fuzzy Euclidean distance between the reference and detected edge images.
Main Results:
- The proposed fuzzy index (FI) was compared against widely used metrics.
- Evaluations were conducted using synthetic images, demonstrating favorable results for the fuzzy index.
Conclusions:
- The fuzzy index (FI) provides an effective alternative for edge evaluation in digital images.
- By avoiding binarization, the fuzzy index preserves information and offers improved evaluation capabilities.
