GSoC 2015: Video Tracking & Image Enhancement/Background Suppression

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

GSoC 2015: Video Tracking & Image Enhancement/Background Suppression

Tom Van den Eynde

Dear all


My name is Tom Van den Eynde, I’m working towards a Master of Science degree in Electrical Engineering at the University of Ghent in Belgium combined with an academic minor in Photonics.  During the first three years of my Bachelor of Science several programming languages were covered. I am familiar and have practical development experience with Matlab, C++ and Python.

Is one of these languages preferred above the others? I think it would be possible to work parallel on this project in two or three of the languages during the GSoC period to give the project extra potential for growth afterwards. Do you think this approach would be beneficial?

Both in my spare time and during curricular projects, I've succesfully implemented a couple of object tracking, color tracking, IR tracking, .. applications. Therefore, I am interested in the suggested project "Video Tracking". 

OpenCV and several others offer extensive examples of object/face/person tracking [1][2][3], therefore designing or implementing such an algorithm shouldn't take up the entire GSoC period. I would like to extend this proposal with real-time video drone steering. Based on object or face detection, an algorithm will determine an optimal path for the drone to follow not to lose track of the object/person. Several practical application come to mind. 
 - The drone could follow a specific person autonomously, either to aid this person in case of emergency or to follow a suspicious person. 
 - The drone could be configured in a search mode, where it scans a specified area for a specific person and can track this person autonomously when found. Several of these inexpensive drones can search large areas very efficiently.

I am also interested in the project "Image Enhancement/Background Suppression". As a first step, I suggest to implement the Drizzle algorithm as used in the Hubble telescope. I think it is feasible to implement this algorithm in the given GSoC period. Any extra time could be used to implement the first steps towards  a "Super-Drizzle" algorithm, which uses adaptive kernel regression.[4] I am willing to implement this second algorithm after GSoC and am looking forward to how it performs compared to Drizzle.

I'd be interested to hear your thoughts about my proposals for these projects and its feasibility. Please feel free to contact me if you have any further questions.

Sincerely,
Tom Van den Eynde
Stud. nr. 01105946
1st Master Electrical Engineering
University of Ghent


------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Opticks-devs mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/opticks-devs
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: GSoC 2015: Video Tracking & Image Enhancement/Background Suppression

tclarke
Administrator

Sorry for the delay responding.

 

Python and Matlab are supported as scripting languages with a limited datacentric API so you can’t easily create new GUI elements, etc. They are great for prototyping and similar activities. C++ is preferred for production quality plugins or plugins requiring an extensive GUI.

 

While your drone steering idea is interesting, I’m not sure it’s really the best application of Opticks which is not really suited for running as an embedded application. It’s designed more for forensic analysis of data and not as much for real-time work. It requires random access to all the data frames and while we’ve been able to feed real-time streaming data into Opticks it’s not really the way it is designed to run.

 

The drizzle idea is definitely worth pursuing. I’d like to see more information on the data you would be targeting. Would you intend to use astronomical images over time like the original Hubble implementation or would you implement drizzle for video data?

 

----------------------

Trevor R.H. Clarke

Software Engineer, Ball Aerospace

(937)320-7087

 

From: Tom Van den Eynde [mailto:[hidden email]]
Sent: Monday, March 16, 2015 1:13 PM
To: [hidden email]
Subject: [Opticks-devs] GSoC 2015: Video Tracking & Image Enhancement/Background Suppression

 

Dear all

 

My name is Tom Van den Eynde, I’m working towards a Master of Science degree in Electrical Engineering at the University of Ghent in Belgium combined with an academic minor in Photonics.  During the first three years of my Bachelor of Science several programming languages were covered. I am familiar and have practical development experience with Matlab, C++ and Python.

 

Is one of these languages preferred above the others? I think it would be possible to work parallel on this project in two or three of the languages during the GSoC period to give the project extra potential for growth afterwards. Do you think this approach would be beneficial?

 

Both in my spare time and during curricular projects, I've succesfully implemented a couple of object tracking, color tracking, IR tracking, .. applications. Therefore, I am interested in the suggested project "Video Tracking". 

 

OpenCV and several others offer extensive examples of object/face/person tracking [1][2][3], therefore designing or implementing such an algorithm shouldn't take up the entire GSoC period. I would like to extend this proposal with real-time video drone steering. Based on object or face detection, an algorithm will determine an optimal path for the drone to follow not to lose track of the object/person. Several practical application come to mind. 

 - The drone could follow a specific person autonomously, either to aid this person in case of emergency or to follow a suspicious person. 

 - The drone could be configured in a search mode, where it scans a specified area for a specific person and can track this person autonomously when found. Several of these inexpensive drones can search large areas very efficiently.

 

I am also interested in the project "Image Enhancement/Background Suppression". As a first step, I suggest to implement the Drizzle algorithm as used in the Hubble telescope. I think it is feasible to implement this algorithm in the given GSoC period. Any extra time could be used to implement the first steps towards  a "Super-Drizzle" algorithm, which uses adaptive kernel regression.[4] I am willing to implement this second algorithm after GSoC and am looking forward to how it performs compared to Drizzle.

 

I'd be interested to hear your thoughts about my proposals for these projects and its feasibility. Please feel free to contact me if you have any further questions.

 

Sincerely,

Tom Van den Eynde

Stud. nr. 01105946

1st Master Electrical Engineering

University of Ghent

 



This message and any enclosures are intended only for the addressee. Please
notify the sender by email if you are not the intended recipient. If you are
not the intended recipient, you may not use, copy, disclose, or distribute this
message or its contents or enclosures to any other person and any such actions
may be unlawful. Ball reserves the right to monitor and review all messages
and enclosures sent to or from this email address.

------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Opticks-devs mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/opticks-devs
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: GSoC 2015: Video Tracking & Image Enhancement/Background Suppression

Tom Van den Eynde
Dear Mr. Clarke

My apologies for the delayed response.

I've done some additional research concerning the usable data for the image enhancement/background suppression project. I believe the use of astronomical images over time is the best choice for the initial phase. Astronomy offers a wide variety of undersampled images over time of the same object. Enormous publicly available databases [1] offer collections of images to test and improve early implementations of the Drizzle algorithm. Moreover, using original images from the WFPC2 camera used by the Hubble telescope would honor the astronomical significance of the Hubble telescope.

Implementing the Drizzle algorithm for video data would be a great follow-up on the project. Once the program works fine for a wide range of astronomical images, I think the expansion to video data would be feasible during the GSoC period.

Sincerely,
Tom Van den Eynde
Stud. nr. 01105946
1st Master Electrical Engineering
University of Ghent


2015-03-17 14:53 GMT+01:00 Clarke, Trevor <[hidden email]>:

Sorry for the delay responding.

 

Python and Matlab are supported as scripting languages with a limited datacentric API so you can’t easily create new GUI elements, etc. They are great for prototyping and similar activities. C++ is preferred for production quality plugins or plugins requiring an extensive GUI.

 

While your drone steering idea is interesting, I’m not sure it’s really the best application of Opticks which is not really suited for running as an embedded application. It’s designed more for forensic analysis of data and not as much for real-time work. It requires random access to all the data frames and while we’ve been able to feed real-time streaming data into Opticks it’s not really the way it is designed to run.

 

The drizzle idea is definitely worth pursuing. I’d like to see more information on the data you would be targeting. Would you intend to use astronomical images over time like the original Hubble implementation or would you implement drizzle for video data?

 

----------------------

Trevor R.H. Clarke

Software Engineer, Ball Aerospace

<a href="tel:%28937%29320-7087" value="+19373207087" target="_blank">(937)320-7087

 

From: Tom Van den Eynde [mailto:[hidden email]]
Sent: Monday, March 16, 2015 1:13 PM
To: [hidden email]
Subject: [Opticks-devs] GSoC 2015: Video Tracking & Image Enhancement/Background Suppression

 

Dear all

 

My name is Tom Van den Eynde, I’m working towards a Master of Science degree in Electrical Engineering at the University of Ghent in Belgium combined with an academic minor in Photonics.  During the first three years of my Bachelor of Science several programming languages were covered. I am familiar and have practical development experience with Matlab, C++ and Python.

 

Is one of these languages preferred above the others? I think it would be possible to work parallel on this project in two or three of the languages during the GSoC period to give the project extra potential for growth afterwards. Do you think this approach would be beneficial?

 

Both in my spare time and during curricular projects, I've succesfully implemented a couple of object tracking, color tracking, IR tracking, .. applications. Therefore, I am interested in the suggested project "Video Tracking". 

 

OpenCV and several others offer extensive examples of object/face/person tracking [1][2][3], therefore designing or implementing such an algorithm shouldn't take up the entire GSoC period. I would like to extend this proposal with real-time video drone steering. Based on object or face detection, an algorithm will determine an optimal path for the drone to follow not to lose track of the object/person. Several practical application come to mind. 

 - The drone could follow a specific person autonomously, either to aid this person in case of emergency or to follow a suspicious person. 

 - The drone could be configured in a search mode, where it scans a specified area for a specific person and can track this person autonomously when found. Several of these inexpensive drones can search large areas very efficiently.

 

I am also interested in the project "Image Enhancement/Background Suppression". As a first step, I suggest to implement the Drizzle algorithm as used in the Hubble telescope. I think it is feasible to implement this algorithm in the given GSoC period. Any extra time could be used to implement the first steps towards  a "Super-Drizzle" algorithm, which uses adaptive kernel regression.[4] I am willing to implement this second algorithm after GSoC and am looking forward to how it performs compared to Drizzle.

 

I'd be interested to hear your thoughts about my proposals for these projects and its feasibility. Please feel free to contact me if you have any further questions.

 

Sincerely,

Tom Van den Eynde

Stud. nr. 01105946

1st Master Electrical Engineering

University of Ghent

 



This message and any enclosures are intended only for the addressee. Please
notify the sender by email if you are not the intended recipient. If you are
not the intended recipient, you may not use, copy, disclose, or distribute this
message or its contents or enclosures to any other person and any such actions
may be unlawful. Ball reserves the right to monitor and review all messages
and enclosures sent to or from this email address.

------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Opticks-devs mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/opticks-devs



------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Opticks-devs mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/opticks-devs
Loading...