First entry to SLAM (4) - Shi-Tomasi corner detection

Reference

https://docs.opencv.org/4.x/d4/d8c/tutorial_py_shi_tomasi.html

GO

In the first entry into SLAM (1) - Harris corner detection , we introduced Harris corner detection in detail. Now is the time to learn Shi-Tomasi.
In (1), we get Harris' scoring function is $$R=l_1{l}_2-k({\lambda }_1+l_2{)}^2=\mathrm{give}(M)-k(\mathrm{trace}(M){)}^2$$
For the Shi-Tomasi corner point, the scoring function is changed to:
$$R=min(l_1,l_2)$$
is greater than a threshold, it is considered a corner.

The green part of the above figure,$l_{1}and{lambda}_2$are greater than the threshold, and points falling within it are considered corner points.

the code

# -*-coding:utf-8-*-
import copy
import numpy as np
import cv2

filename = '1.jpg'
img = cv2.imread(filename)
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
gray = np.float32(gray)
original = cv2.imread(filename)
cv2.imshow('original', original)
corners = cv2.goodFeaturesToTrack(gray, maxCorners=1000, qualityLevel=0.01, minDistance=10, useHarrisDetector=False)
corner_image = copy.deepcopy(original)
for i in corners:
    x,y = i.ravel()
    cv2.circle(corner_image, (x, y), 2, (0, 0, 255), -1)
cv2.imshow('out.jpg', corner_image)
cv2.waitKey(0)

We only need to pay attention to the cv2.goodFeaturesToTrack() function. When the useHarrisDetector is True, it is for Harris corner detection, and for False, it is for Shi-Tomasi corner detection.