A collection of tree simple python scripts that helps finding logos/watermarks on a video file using opencv library. Created for learning pourposes in order to try to clasify some videos based on the producer.
Used to create a bitmap image of a single frame of the video.
- videoPath = Full Path of the video
- filename = Output filename of the image.
- frame = Frame Number to process
Used to get to top/left/treshold position of the logo on the specified video frame. This will be used in order to avoid false positives on the final detection.
- templatePath = Path of the image/logo to find on the frame
- videoPath = Full Path of the video
- frame = Frame Number to process
Used to process the video file and find the logo on a range of frames. Uses opencv matchTemplate method.
- templatePath = Path of the image/logo to find on the video
- videoPath = Full Path of the video
- width = Horizontal Size of the video in pixels
- height = Vertical Size of the video in pixels
- initFrames = Initial frame to start looking for the logo
- maxFrame = Final fram to end looking for the logo
- top = Expected Top Position of the logo (provided by getTopLeft.py)
- left = Expected Left Position of the logo (provided by getTopLeft.py)