range selection
 - <a href="/en/AI%20Cluster%20Description">AI Cluster Description</a>
<img src="https://gyazo.com/437c726a381131c4f23ca322691ceb0c/thumb/1000" alt="image">
Hmmm, a rectangular selection of connected components would cause overlap, that&#39;s true...
Another approach is to divide every 10 squares.
<ul>
<li>That would eliminate the crossing problem, and it wouldn&#39;t look messy.</li>
<li>It happens that &quot;natural clusters&quot; get split up.</li>
</ul>
In the use case I&#39;m using Kozaneba, a human being is doing the collecting of similar things in the first place.
&quot;I want to put this in one group, this is not.&quot;
Because the gap is driven by the decision to
These overlap problems don&#39;t happen.
<ul>
<li>I knew the route and what to do with more than 1,000 cases,</li>
<li>Kozaneba is structured while creating less than 1,000 stickies.</li>
<li>I think it&#39;s better not to equate them indiscriminately.</li>
</ul>
Merging and squaring of overlapping objects
<ul>
<li><img src="https://gyazo.com/1992204dfafe40f76cd8c0adab2f84f2/thumb/1000" alt="image"></li>
<li>Only merging if there is more than 30% overlap now.</li>
<li>This is what happens when you reduce this to 20%.</li>
<li><img src="https://gyazo.com/403217eebe831a46dd705e26132b4691/thumb/1000" alt="image"></li>
<li>Maybe it&#39;s a chain of connections around here.<ul>
<li><img src="https://gyazo.com/23ef63b297d7e1ef990c26a536dd8f8a/thumb/1000" alt="image"></li>
</ul>
</li>
</ul>
<hr>
This page is auto-translated from <a href="https://scrapbox.io/nishio/pKozaneba2025-08-29">/nishio/pKozaneba2025-08-29</a> using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at <a href="https://twitter.com/nishio_en">@nishio_en</a>. I&#39;m very happy to spread my thought to non-Japanese readers.