Image segmentation

Question 1

types of image segmentation?

Answer 1

• By connecting detected edges

- By grouping an image into separate region by area or distinct traits

Question 2

Image contouring

Answer 2

• Continuous curves that follow edges along a boundary
• Into about the shape of an object
• In open cv contours are best detected with white objects against a black background

Question 3

important functions

Answer 3

• cv2.findContours
• cv2.drawContours
• fitEllipse
• boundingRect
• rectangle
• HoughLinesP

Question 4

contour features

Answer 4

• Every contour has a number of features that you can calculate, including area, orientation (the direction that most of the contour is pointing in), it’s perimeter,
• http://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_imgproc/py_contours/py_contour_properties/py_contour_properties.html

Question 5

Hough space

Answer 5

• Hough transform converts a line in image space to a point in hough space. Each point could be the slope m and the intercept b o r (m,b)
• You can also use polar coordinates. is the perpendicular distante from the origin to the closest point in the line and the angle from this point
• the intersection of lines in hough space is a line in the image space and suggests a detected edge or boundary

Question 6

kmeans clustering

Answer 6

• Divide image in segments by clustering data points with similar traits
• Unsupervised learning. Machine learning no data labeling
  1) Choose k random center points
  2) Assign each data point to its nearest center point
  3) Take the mean of each cluster. The mean becomes the center point
  4) repeat 2 3 until convergence