Ask Your Question
0

OpenCV for Secret Agents!

asked 2013-08-28 06:16:35 -0600

SamB gravatar image

updated 2013-08-28 06:18:45 -0600

Hi folks!

I'm Sam Birch, a commissioning editor at Packt (www.packtpub.com). We are going to produce a book with the above title - and yes, we want it to be as incredible as that sounds! Possible projects include...

  • a robot that can acquire a specific target and follow it, whilst recording what the target does
  • a lie detector using motion tracking to measure pulse rate (like a cheap, non-invasive polygraph)
  • a face recognition doorway entrance system (useful for that volcano-top secret lair)

...along with loads of other cool ideas. I'm interested in talking to anyone who either a) is interested in writing the book, or b) has any fantastic suggestions for projects to include (or c) both, which is preferable!). These are only ideas, if you're interested in working on this project then we can have a chat about what actual projects we can do! Give me a shout at [email protected]

Cheers,

Sam

edit retag flag offensive close merge delete

Comments

Actually after reading the authors comment I think their main goal is interesting, but I still doubt some of the topics :) And yes, I still agree that working out projects for interesting stuff, that do good, is still a better approach ;) Go cure cancer you!

StevenPuttemans gravatar imageStevenPuttemans ( 2013-08-28 07:28:08 -0600 )edit

The topics are just examples to stimulate people's imaginations, based on real projects that have already been developed. The plan was to use them as a jumping-off point, not a definite 'we will certainly include these'. It's definitely more garden shed than Wayne Enterprises. I'd love to be able to help cure cancer, but I'm not convinced it'd be possible!

SamB gravatar imageSamB ( 2013-08-28 07:37:28 -0600 )edit

1 answer

Sort by ยป oldest newest most voted
4

answered 2013-08-28 06:32:43 -0600

Actually, from my own point of view I think this is the most terrible idea I have heard so far ... Let me explain why. Computer vision libraries like openCV have been introduced at first to stimulate a universal research help and to push people to actually use academic developped techniques into commercial products.

Many of the research in 'secret agent' stuff is funded by military divisions all over the world. One of the main reasons why all of this research stays closed for the public, is the fact that this opens doors to people, not respecting other peoples privacy.

Take for example the employer-employee situation. Since cheap camera systems have been created and sold, almost every employer thinks it is basically his right to spy on his employees. Only look at the amounts of forum questions that handle the detection and tracking of people through rooms, looking at how people behave with face detectors, ...

In my opinion there is a very thin line between what is morally correct and what is morally incorrect, which leads to many solutions being used for the wrong purpose. Stimulating this general behaviour by publishing a book containing projects that can effectively invade anothers persons privacy, seems just plain wrong.

I think there is a need for more interesting books, like:

  • How to apply OpenCV computer vision algorithms in industrial applications
  • How to apply OpenCV computer vision algorithms in medical/healthcare applications
  • How to apply OpenCV computer vision algorithms in food processing
  • ...

Just my two cents :) which are basically an opinion of someone, not more, not less :) If you feel like discussing this problem, feel free to contact me. If you would actually consider publisching a book containing these projects, I think an introductory chapter explaining the regulations of privacy and the influence of computers in a daily environment should be very interesting.

edit flag offensive delete link more

Comments

Hi Steven,

Thanks for your feedback. We don't see this title as anything other than a fun, exciting project book for hobbyists who enjoy tinkering and showing what they've made. There is, to be honest, very little real-world application of, for instance, the robot project idea. It is merely for people to have fun. Also, I suspect the military/government types you allude to are already way ahead technology-wise, plus are unlikely to utilize open source software that is freely available to all. This fun approach for hobbyists has been very popular in the past with, for example, 'Raspberry Pi for Secret Agents' and '30 Arduino Projects for the Evil Genius', neither of which cause concern about serious misuse.

Cheers,

Sam

SamB gravatar imageSamB ( 2013-08-28 06:58:50 -0600 )edit

I do not fear the military/government, I do fear the regular hobbiest, which will just use premade projects to do things he couldn't do before. Thats the whole deal about opensource software. I think there are way more interesting things for hobby focused people than 'secret agent'. How about 3D reconstruction, robot targetting (only one I agree upon), measuring objects from images, applying image filters to create photoshop like effects, ...

StevenPuttemans gravatar imageStevenPuttemans ( 2013-08-28 07:04:27 -0600 )edit

Also, your comment about very little real-world applications of robot projects, I think you are focussing on the wrong software package. Ever heard of ROS? It can be combined with OpenCV quite easily to do amazing robot stuff. There are thousands of hobbyists already using this approach.

StevenPuttemans gravatar imageStevenPuttemans ( 2013-08-28 07:06:44 -0600 )edit
1

Hi Steve - the projects will require extra stuff - for instance we are looking at using a Raspberry Pi or BeagleBone to run a few of them, plus software such as ROS. The unifying theme would be OpenCV though. We only want people to be limited by their imaginations! (and common human decency of course). I should also point out that this is only one of a series of titles on OpenCV from Packt - this is not our only offering!

SamB gravatar imageSamB ( 2013-08-28 07:12:43 -0600 )edit

Actually from a technical point of view, a project containing a raspberri Pi with ROS and OpenCV working together, would mean that the raspberri pi is the unifying theme :P But I quite like the idea to make a complete project based on several techniques. For that I would actually be interested! However, I would then suggest taking 1 large project and use the subsets as chapters like for example, chapter raspberri pi, chapter opencv, chapter ros, chapter that combines all, ...

StevenPuttemans gravatar imageStevenPuttemans ( 2013-08-28 07:23:56 -0600 )edit

I see what you mean with regards to that project. I meant that it would be the unifying theme between the separate projects, as some would not require a Raspberry Pi. The large single-project idea was mooted, but we decided that 6 to 8 individual projects would give a reader more scope to use the information as a template that they could then play around with, rather than a hard-and-fast blueprint.

SamB gravatar imageSamB ( 2013-08-28 07:43:12 -0600 )edit

Though challenge I think :) Expecially if you want all projects to be described evenly well, you will first need people to actually develop the projects, unless you already have hand made projects available.

StevenPuttemans gravatar imageStevenPuttemans ( 2013-08-28 07:46:39 -0600 )edit
1

We don't have them pre-made, and that is a difficulty with a title like this. One possible approach would be a group of authors, all of whom contribute a project or two.

SamB gravatar imageSamB ( 2013-08-28 08:06:11 -0600 )edit

Question Tools

Stats

Asked: 2013-08-28 06:16:35 -0600

Seen: 1,773 times

Last updated: Aug 28 '13