Program Idea: Enforced Dressing Program

[TwistedGenius]

First up, I'm not a programmer. I know all the programmers who are around are quite busy, but I believe the idea I've got fills a need currently lacking among the suite of programs we've currently have access too.

Currently we have random dresser programs but no way to enforce the wearing of outfits. I believe I have found a way to do this (conceptually) but I need a programmer to tell me whether it is possible, and then (should they be so kind) to create it.

Summary:

Using a webcam, the program takes a picture of the individual standing before it. If the image taken matches a reference picture (taken beforehand) of the individual wearing a desired outfit (within a certain tolerance) then the program registers a match (using image comparison software - similar to Reverse Google Image Search), and runs a pre-selected release method (or program).

Suggested Layout:

The program would need rather few features:

- Take reference picture (button): After a 5 second delay (to give time to get into position), takes a picture and saves it as the reference picture to be used in later comparisons.
- Load reference picture (button): Alternatively a photo could be taken beforehand (or load up a reference picture from a previous session).
- Check outfit (button): After a 5 second delay (to get into position), takes a picture and then compares it to the reference picture. Alternatively, it could just check every minute - this would avoid having to click a button if restraints are part of the enforced dressing.
- Release (button): Select release program or opening the CD tray (possibly with a delay if you like).
- Viewing window: Webcam feed so you can make sure you're in the same picture/frame as the reference picture
- Reference picture: Shows the reference picture so you know what you.


Example Scenario:

Obviously if you're getting dressed then there's a chance you're not wearing anything you're being tied to requiring release (yet), but it could work quite well in conjunction with other programs we already have. As an example:

You have two laptops/computers (A and B). Computer A is running the Enforced Dressing Program (EDP) and B is running your release program of choice (Blowjob Trainer, Motion Detector, Memory Game etc). Now, imagine you also have a chastity device and some manner of lockable clothing:

1. Setup:
- Set up the key to the chastity device to Computer A (running EDP), and have the reference picture already set for the outfit you'll be forced to wear.
- Set up your key/release for the clothing/restraints (that you'll be forced to put on) to Computer B (running whatever release program you want).
- Put on the chastity device

2. Now in order to get the key to the chastity device, you're going to have to get dressed as the EDP is set to. The program won't release the key to the chastity device until you match the clothing/pose of the reference picture. This could be a maid's outfit, certain restraints (or both).

3. Once you're dressed as required, click the button (or wait for the next scan if just set to check every minute) and if it matches then you get the key to the chastity device! But wait... now you're dressed and the clothes are locked on

4. Now you've got to go and do whatever Computer B requires you to do in order to get the keys to your clothes/restraints before you can make use of the key to your chastity device. Being forced to sit still or suck a dildo because a computer is making you do it just because more humiliating because now you're doing it dressed however another computer wanted you to.

Thoughts?

The release code (that is the code which would control key release or executing another program) for this program is kicking around in most of the programs that are already done, as is the part where it takes the photos for the comparison (I know the blowjob trainer takes photos). Aside from the GUI, I figure the most work will be coming up with the image comparison coding and working out what is a practical level of difference between the two photos.

I originally thought this could be done using the Color Detector program, but I figured it could be too easily beaten (just hold up the clothing to the camera rather than wearing it) - I wanted something a little harder to lie to.

I look forward to hearing your ideas on this.

[LoKiT]

Hi TwistedGenius, I like the idea. Technology is rapidly progressing for image recognition and there are some good examples about.
From personal experience and playing with this software you soon figure there are problems that creep into the equation. Like light levels and angels and camera noise.
Quality and performance is getting better and I have no doubt we will see it soon one day soon in game play

[sophiestacey]

This would be pretty hard to code. Whats stopping the user from just holding up i.e. the dress to the camera and it registering it? The whole image recognition would not only need to detect that it is the correct outfit they should be wearing but that it is actually on their body. In theory the user could cheat and use a real life manikin type frame to make it look like they are wearing the outfit. So it would also need to detect that it is *you* and not a dummy.

Image recognition is one of the most complex tasks in computer science and although it seems easy for us humans to detect, to teach a computer requires advanced mathematics and detection algorithms.

It'll probably be quicker and more fun to find a mistress/master to train you

[Jadit]

There's many algorithms for camera software, but i would edit the camera image, sort of filter before reading anything from it.
- Resize image smaller, if camera data is too much. Something like a shirt can be like 50000 pixels in 1 normal sized image. You don't need that much, and it will make reading faster. Use linear-filtering or something for scaling so it averages out the "random"-like colors.
- Min-max the brightness. Find brightest and dimmest pixel in the camera, and scale everything to that range. (Optional)
- RGB color channel averaging. For example average every color to its nearest of 0, 127 or 255. If it's too rough, try steps of 64.

Now you have image that is much easier to get readings from. Should only take fraction of a second to process it all.

[TwistedGenius]

In theory the user could cheat and use a real life manikin type frame to make it look like they are wearing the outfit. So it would also need to detect that it is *you* and not a dummy.

If you've gone to enough effort to acquire a dummy and dress it in the clothes to beat the program, there's a good chance you didn't want to be using it in the first place. There's a certain willingness to all bondage scenarios unless someone is in the room with you and able to place you under duress.

There's many algorithms for camera software, but i would edit the camera image, sort of filter before reading anything from it.
- Resize image smaller, if camera data is too much. Something like a shirt can be like 50000 pixels in 1 normal sized image. You don't need that much, and it will make reading faster. Use linear-filtering or something for scaling so it averages out the "random"-like colors.
- Min-max the brightness. Find brightest and dimmest pixel in the camera, and scale everything to that range. (Optional)
- RGB color channel averaging. For example average every color to its nearest of 0, 127 or 255. If it's too rough, try steps of 64.

Yeah, that's the sort of thing I was thinking of (I'm just not a programmer)- as long as the colours and shapes are in the right places it unlocks. As long as it can't easily be beaten by just holding up the outfit (which would be difficult anyway if it involved more than one piece), it's doing it's job. You'd never get the exact photo twice, so toning tollerance/resolution it down would be fine.

[sophiestacey]

In theory the user could cheat and use a real life manikin type frame to make it look like they are wearing the outfit. So it would also need to detect that it is *you* and not a dummy.

If you've gone to enough effort to acquire a dummy and dress it in the clothes to beat the program, there's a good chance you didn't want to be using it in the first place. There's a certain willingness to all bondage scenarios unless someone is in the room with you and able to place you under duress.

I was more making the point that such a crudely implemented program can be easily fooled even creating a dummy-like figure, amongst lots of other cavets. There is a "self bondage blowjob trainer" on this site which uses colour matching and it's pretty unreliable and easily fooled. Ideally, image recognition should be used to detect when sucking etc but this is much harder than it seems and takes a lot of computational power. If it were easy, it would have been done. Doing a full dress up enforced program would be much, much harder than a blowjob trainer program unfortunately.

Of course all bondage scenarioes require a "will" but to go to the effort to create a program that can be pretty easily be fooled seems like a lot of effort and waste of time. Why not just use that will power to dress and stay dressed?

Soph