Computer vision Jobs

5 were found based on your criteria

  • Fixed-Price – Est. Budget: $1,000.00 Posted
    We need ability to get information of all textblock and image boundary coordinates of ePaper pages (converted into 24Bit/300dpi TIFF's). All correlative text must be connected as rectangular or polygonal boundary. Also invers Text or Text on colored Background must be detected. No OCR is needed! It is absolut essential to accomplish the detection in less than 1,5 seconds/page (based on pagesize of approx. 4000x5200 pixel) Attached is a sample of the results we expect. The ...
  • Fixed-Price – Est. Budget: $80.00 Posted
    I am looking for a developer who can help me with part of a project. I need the ability to pass an image to my server from special paper and recognize defined text on a sign and run a function based on what was scanned in. You must be willing to sign a NDA to work on this project. Please apply for more information.
  • Fixed-Price – Est. Budget: $500.00 Posted
    I'd like to extract TV commercials from live stream of video file to a series of video files. The use of Comskip (https://github.com/Rupan/comskip) seems to be the good option, or any other one. The idea of this is to create a video database of TV commercials. Input could be "split.py test_video" And the output: test_video1.mpg (commercial 1) test_video2.mpg (commercial 2) test_video3.mpg (commercial 3) etc.. Programming language can be c/c++, or ...
  • Fixed-Price – Est. Budget: $2,000.00 Posted
    Good morning I would meet a high skilled in computer vision developer to build a method to find and track patterns ( brand logos ) in video files. The system should be able to detect the position of every logo position within the files and should be able to detect the percentage of visibility in case of occlusion. The system should be as fast and accurate as possible, OpenCV version 3 and GPU used should be used. Algorithm will be deployed in ...
  • Hourly – Less than 1 week – Less than 10 hrs/week – Posted
    I would like to find someone that can build an efficient OpenCV based algorithm takes as input a high resolution camera (iPhone) image of a projected chart (projected by Powerpoint onto a screen) and compares it to the actual slides (taken from a PDF generated from Powerpoint) and returns the matching image, or nothing if no image matches. It doesn’t need to be real time, but it does need to run on an iPhone. I have built the framework ...
loading