Edge_Detection

Classical edge-detection algorithms using python

1. Edge-detection operators based on first derivative

• Roberts.py
• Prewitt.py and Scharr.py
• Kirsch.py and Robinson.py —— Update: Robinson operator included in Kirsch.py
• Sobel.py and Sobel_Normalize.py —— good algorithm
• Canny.py —— best algorithm. We can assume that the problem of edge-detection has been solved by this algorithm.

Tip1: Algorithms using first-derivative information are to detect the positions of gray-value change in the image.

2. Edge-detection operators based on second derivative

• Laplacian.py
• LoG.py

Tip2：Various algorithms (just list 2 algorithms here) that use the second-derivative information are not designed to exceed the above algorithms (first-derivative), but rather they can detect some special boundary —— The positions of the pixel where the gray-value changes the fastest!

Necessary python packages

• Numpy —— pip install numpy
• Scipy —— pip install scipy
• opencv —— pip install opencv-python

Tip3: python > 3.6. Codes are best opened and run with PyCharm IDE.

---

Update1:

• Sobel_Normalize.py
• LoG_New.py