Image stitching

Hello. I'm Ph.D. in image processing with 15 papers in the relevant journals. I'm the best mathematician here and all my projects were completed with the best rates. Also i know about quadracopters and its applications. I definitely can help you. Thanks for consideration.

Hi! I am professional C/C++/C#/Java programmer, I can do this project with highest quality. Best Regards, Szymszteinsl

Hello, I represent a small team of software developers from Ukraine that is interested to be involved in your project. At Jazzros we are involved in development of sophisticated solutions like image processing software, CAD solutions etc. Our engineers have strong mathematics background that helps us to resolve difficult engineering problems. Regarding your project. We already have some things to start this project. In general it will based on Neural Networks(NN) and Principal Component Analysis(PCA). But degree of complexity depends on real data which will be in processing. So, please provide me source images which should be connected. As more as possible. For example, NN can be useful for combining images when they have common intersection section at the edge. If they will not have such sections NN cannot help. PCA will help to calculate movement-vector to combine two connected images to make one. One more question - how do you plan to check the degree of completion of work? Best Regards, Borys

hi, I can get this done easily using opencv or matlab .By using various image processing techniques like edge detection and feature matching to conect images together,i can make a final good quality jpg file. Also I have done connecting upto tw0 images in matlab which can be upgraded to ten. Can u send me some sample files? hope to hearing fromu soon. thankz

Hello Jonathan, I have read your description and clarification message, and am interested in the project. Recently I have finished my PhD in CS/AI, and am running a world class robotic football team. During these projects, as well as during my university time before these, I have been very much interested and involved in computer vision, and I enjoy working in this field a lot. Actually, I have now been hired in a position where one of my tasks will be very close to your project: combining aerial photos into one overview. Also in my track record here at Freelancer.com you can see I have successfully undertaken several image processing projects. Another relevant project is the one in which I created a program that stitched together 3D point clouds obtained with multiple passes of a laser scanner; the algorithm I used there was an adaptation of one originally developed to stitch photographs together. Could you tell me a bit more about the project? Most importantly, will the photos be made beforehand and then loaded into the program, or will the program take images live from the Quadcopter? Also, is there possibly any metadata available from the Quadcopter, such as GPS, or sensor readings from between the images were taken? And finally, what kind of platform will the program run on? Will it be Windows, OSX, Linux, or should it preferably run on any of these? None of this should be a problem, I am just curious. Looking forward to hearing fro you in the near future. Sander

Hello. I am a programmer and computer scientists focused on photogrammetrical modeling of "virtual cities" (we built virtual cities from aerial/satellite/ground photographs). At our local university I created course about computational photography, where your subject belongs to. Here in Europe, I am cooperating with Eurosense (photogrammetry producer for GIS), Vexcel (innovative aerial cameras from prof. Leberl) and in the past with CyberCity AG (one of the first software for building 3D cities from aerial photos). Solving of your problem depends on the input. It can range from using ready-made commercial software, e.g. Photomodeler (needs interactive user input), adapt existing open solutions like Hugin (GPL license), use well-known computer vision libraries for automatization (OpenCV), to the most tedious work - implement own algorithms. At first, there will be couple of questions about the level of automatization of the whole process, quality of cameras (options from calibrated profi systems to ...cameras with unknown parameters to developer) and the environment (no one can guarantee automatic uncalibrated photo-stitching of homogenous environments like rain forest taken from the air with 100% precise results).

Hello, We are team consisting of two software engineers experienced in computer vision and algorithms. We are able to solve this problem, but we need more information. Please send us a private message. For example whether pictures are taken from the fixed positions. Example photos would be great.

I am ready to provide you this solution. As we disccus, for me is clear and we can move forward this project as well

A proposal has not yet been provided

I'm an electrical engineer with 6+ years of image processing and remote sensing experience using C/C++. I can easily do this for you using OpenCV.

We are focused on large projects, currently we work with VW Frankfurt , doing some complex applications as an addon to Citrix and their VPN's > Reference can be given without any problem. I would be happy to discuss this in details and why not to start a long term business commitment. Thank you, Bogdan

Hello, I can help. I have the necessary imaging background to complete the task. 10 images stitched into a single image. Only possible options are 5x2 or 2x5 or 10x1 or 1x10. Edges will be matched up to give a final image. Let me know if you do need my services. Best regards, AccuPro - Providing simple solutions to complex problems!

Hi We have experience and expertise of image processing applications. We have been working on couple of image processing applications right now in the field of medical imaging. Kindly contact us and we can answer all your questions and would love to do this project for you. Regards Imran

Dear Mr./Ms. I am Matías Santiago Díaz, an Electronic Engineer graduated at the Universidad Católica de Córdoba, Argentina. In Argentina I worked in a company that provides technological solutions. I have experience in C/C++ and Java programming; Experience in Linux environment programming, using tools like Eclipse, Qt Creator, CrossWorks, SVN, Mantis BugTracker; Experience in embedded systems programming, real time operating systems, and Linux Embedded. Best regards, Matías Díaz.

I've almost 6 years of experience in doing the computer vision projects notably, "Passive Ranging(absolute depth perception) using Stereo Vision", "Robust Bullet Marks detection in video sequences, for Fast scoring in Closed Target Ranges"... My experience in the solution of correspondence problem in stereo vision will prove to be quite handy in your project, IA. Following are the basic steps in developing panoramas: STEP-1: Determine the appropriate motion model relating pixel coordinates in one image to pixel coordinates in another. STEP-2: Estimate the correct alignments relating various pairs of images, using either direct pixel-to-pixel comparisons combined with gradient descent or feature-based alignment techniques. STEP-3: Developing algorithm to compute globally consistent alignments from large collections of overlapping photos. STEP-4: Once the alignments have been estimated, step will be to choose a final compositing surface onto which to warp and place all of the aligned images. STEP-5: Seamlessly blending overlapping images, even in the presence of parallax, lens distortion, scene motion, and exposure differences. Kindly give me details of your application so that I may be of assistance to you. You can hire me for effective help, IA... Regards, Imagingscience