ImageEn for Delphi and C++ Builder ImageEn for Delphi and C++ Builder

 

ImageEn Forum
Profile    Join    Active Topics    Forum FAQ    Search this forumSearch
 All Forums
 ImageEn Library for Delphi, C++ and .Net
 ImageEn and IEvolution Support Forum
 Proposal for a new Magic Wand Selection algorithm

Note: You must be registered in order to post a reply.
To register, click here. Registration is FREE!

View 
UserName:
Password:
Format  Bold Italicized Underline  Align Left Centered Align Right  Horizontal Rule  Insert Hyperlink   Browse for an image to attach to your post Browse for a zip to attach to your post Insert Code  Insert Quote Insert List
   
Message 

 

Emoji
Smile [:)] Big Smile [:D] Cool [8D] Blush [:I]
Tongue [:P] Evil [):] Wink [;)] Black Eye [B)]
Frown [:(] Shocked [:0] Angry [:(!] Sleepy [|)]
Kisses [:X] Approve [^] Disapprove [V] Question [?]

 
Check here to subscribe to this topic.
   

T O P I C    R E V I E W
PeterPanino Posted - Aug 29 2017 : 13:49:55
I would like to propose a new Magic Wand Selection method, as I find the current Magic Wand Selection action very inefficient:

1. While holding down a modifier key, clicking on several positions collects the colors of the clicked pixels in a ColorList. Then the lifting-up of the modifier key ends the collecting of colors.

2. Then, when clicking on any pixel position, does not use only this clicked pixel's color for the Magic Wand processing but all the colors previously collected. This would be much more efficient.

It could also be combined with an option where in step 2 not a single click starts the Magic Wand process but a brush-painting mouse tool is dragged and is effective only on those pixels under the dragged brush tool which correspond to the previously selected colors.

What do you think?
4   L A T E S T    R E P L I E S    (Newest First)
xequte Posted - Aug 31 2017 : 00:42:09
Hi

We have implemented a new method that allows selection using the ChromaKey algorithm. You can email us to test it, if you like.

Nigel
Xequte Software
www.xequte.com
nigel@xequte.com
xequte Posted - Aug 30 2017 : 19:41:19
> So maybe a variable hotspot size could be something to try.

Yes, that would be cool, especially if the cursor enlarged to match the hotspot size. I'll look into this for a later update.



Nigel
Xequte Software
www.xequte.com
nigel@xequte.com
PeterPanino Posted - Aug 30 2017 : 19:18:50
Hi Nigel,

I will try Chromakey tomorrow.

I once made some experiments. For example:

While still holding down a modifier key, each clicking increases the tolerance. In this way, you could expand the selection area by just clicking repeatedly on the same place with the CTRL key pressed. Then, when lifting the CTRL key the tolerance goes back to standard value and you can play the same game on a place nearby which is not already selected. This works pretty well.

The main problem with the current MagicWand selection: No matter how exact the user tries to be there always remain small hotspots of unselected space between the existing selections. These unselected hotspots are very difficult to hit, so it is difficult to create a contiguous area of selection. So maybe a variable hotspot size could be something to try.

Just the other day an idea similar to your chromakey thing came to my mind and it's funny that you just mention it.
xequte Posted - Aug 30 2017 : 18:07:27
Hi Peter

We have considered making another magic wand mode that grabs more colors, in a similar way to ChromaKey:

https://www.imageen.com/help/TImageEnProc.RemoveChromaKey.html

(If you try the ChromaKey demo, you will see that it can use all colors in a selection).


The hard bit is adding it in a way that is practical for users to access. I like the idea of being able to paint an area for removal.


In the meantime, have you tried increasing the tolerance?

https://www.imageen.com/help/TImageEnView.MagicWandTolerance.html

Nigel
Xequte Software
www.xequte.com
nigel@xequte.com