Mean shift algorithm #235

ashwani-rathee · 2021-09-19T07:43:32Z

covers mean shift image segmentation algorithm

timholy

I hadn't noticed these (they arrived back when I wasn't paying attention), thanks for adding this!

timholy · 2023-08-02T17:10:28Z

docs/examples/image_segmentation/mean_shift.jl

+# -  We have specify a prior on the number of clusters(Number of K which is no of
+#    segments).


Suggested change

# - We have specify a prior on the number of clusters(Number of K which is no of

# segments).

# - We have specify a prior on the number `K` of clusters (segments).

timholy · 2023-08-02T17:11:49Z

docs/examples/image_segmentation/mean_shift.jl

+# but has two main issues:
+# -  We have specify a prior on the number of clusters(Number of K which is no of
+#    segments).
+# - It can be sensitive to value of k-means parameter.


This seems not entirely surprising. Is it really worth making as a separate point?

It could also be stated that K-means tends to return clusters of approximately equal radius, which may not be how a human would do it.

timholy · 2023-08-02T17:13:08Z

docs/examples/image_segmentation/mean_shift.jl

+# - Each hill represents a cluster
+# - Peak(mode) of hill represents "center" of the cluster
+# - Each pixel climbs the steepest hill within its neighborhood
+# - Pixel assigned to the hill(cluster) it climbs. That's the idea behind meanshift. 


Suggested change

# - Pixel assigned to the hill(cluster) it climbs. That's the idea behind meanshift.

# - A pixel is assigned to the hill(cluster) it climbs.

I think the "idea behind meanshift" encompasses all of these points.

timholy · 2023-08-02T17:13:58Z

docs/examples/image_segmentation/mean_shift.jl

+#   The Manhattan Distance is another useful choice sometimes.
+# - A radius. All pixels within this radius (measured according the above distance) 
+#   will be accounted for the calculation.
+# - A value difference. From all pixels inside radius r, we will take only those 


Would "intensity" be clearer than "value"?

timholy · 2023-08-02T17:14:32Z

docs/examples/image_segmentation/mean_shift.jl

+# - Set $m_i = f_i$ as initial mean fro each pixel idea
+# - Repeat the following for each mean mi:
+#   - place window of size $w$ around $m_i$.
+#   - compute centroid m within the window set $mi = main$


Not sure what main means

timholy · 2023-08-02T17:14:49Z

docs/examples/image_segmentation/mean_shift.jl

+# Task: Find modes(clusters) of distribution. 
+
+# Clustering:
+# - Set $m_i = f_i$ as initial mean fro each pixel idea


Suggested change

# - Set $m_i = f_i$ as initial mean fro each pixel idea

# - Set $m_i = f_i$ as initial mean for each pixel

timholy · 2023-08-02T17:15:12Z

docs/examples/image_segmentation/mean_shift.jl

+
+# Clustering:
+# - Set $m_i = f_i$ as initial mean fro each pixel idea
+# - Repeat the following for each mean mi:


Use the same notation. Is mi the same as $m_i$ ?

timholy · 2023-08-02T17:16:00Z

docs/examples/image_segmentation/mean_shift.jl

+#   - stop if shift in mean $m_i$ is less than a threshoold $\epsilon$, $m_i$ is the mode
+
+# It's a very simple but computationally expensive method which doesn't require input of
+# the number of clusters. Very robust to outliers and clustering depends on window size w.


Technically, one could say that it replaces specification of # of clusters with the size of the radius.

mean shift algorithm

5459af6

ashwani-rathee changed the title ~~[WIP]: Mean shift algorithm~~ Mean shift algorithm Sep 25, 2021

timholy reviewed Aug 2, 2023

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Mean shift algorithm #235

Mean shift algorithm #235

ashwani-rathee commented Sep 19, 2021

timholy left a comment

timholy Aug 2, 2023

timholy Aug 2, 2023

timholy Aug 2, 2023

timholy Aug 2, 2023

timholy Aug 2, 2023

timholy Aug 2, 2023

timholy Aug 2, 2023

timholy Aug 2, 2023

		# - We have specify a prior on the number of clusters(Number of K which is no of
		# segments).

	# - Pixel assigned to the hill(cluster) it climbs. That's the idea behind meanshift.
	# - A pixel is assigned to the hill(cluster) it climbs.

	# - Set $m_i = f_i$ as initial mean fro each pixel idea
	# - Set $m_i = f_i$ as initial mean for each pixel

Mean shift algorithm #235

Are you sure you want to change the base?

Mean shift algorithm #235

Conversation

ashwani-rathee commented Sep 19, 2021

timholy left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment