Hello,
We are trying to make a custom ffmpeg filter which can hide the logo from video input source and put an overlay image or delogo the original logo and hide it dynamically.
Definitions:
------------
- Input source: Which can be a local file or live input stream (UDP, RTMP, HTTP) having the unwanted dynamic logo with random position, time to show, time to hide, colors and background color.
- Logo pattern file: Which is a local file (PNG file) having the pattern of the unwanted logo to be removed.
- Output destination: Which is the final output file, Output stream can be (UDP, RTMP, HTTP) to be exported by ffmpeg after processing the filter and encoding options.
- Similarity rate: Which is the percentage for the similarity between the logo pattern file and input source to make the ffmpeg recognize that unwanted logo from the pattern file.
- Overlay: This option can be a local file (PNG File) to be appended over the original logo or using delogo filter which is blurring and removing the original logo.
Procedure:
----------
- We will provide testing files (MP4) which have the examples for the unwanted logo to be removed.
- We will provide the logo pattern file which holds the logo pattern regardless the logo color and background color and size of the input source logo.
- Logo pattern file and similarity rate should be treated as inline parameters with the ffmpeg complex filter options.
- Example for the complex filter options: # ffmpeg -i .... -filter_complex "removeLogo@logo=/home/test/[login to view URL]@similarity=0.45@overlay=/home/[login to view URL]" -o flv rtmp://x.x.x.x/xx/xx
- We are expecting testing output file.
- Source code should be provided when completing the project.
- Building ffmpeg should include and support the use for the following modules: libx264, x265, libaac, v4l2, alsa, nvenc (NVIDIA GPU Support) , libmfx (Intel QuickSync Support).
- Static build compatible with CentOS 7 and Ubuntu
Hello dear, I have a lot of experience in Computer Vision applications / image and video processing in mobile and windows applications,
My Team experience includes Port Based container Damage detection Software: detecting damaged container and keeping record and report of containers movement, Driver fatigue detection using raspberry pi which in a matter of milli seconds determines if the driver is sleeping or not and alarm accordingly using OpenCV, Vehicle counting in the particular category they belong to on a high speed road, number plate recognition using OCR algorithm of openCV, face detection and recognition for attendance system of a school and patient management system of a medical system.
In short We have all the experience needed to complete the project with maximum accuracy.
I will be waiting for your kind response.
Regards
Hi, Hope you doing well. I read your project description and very confident to
handle this project perfectly as we are a lot expertise in image processing and openCV
Let's have a discussion, negotiate costing, timeline and then we proceed
Thanks!
We have been working with OpenCV and FFMPEG for several years now. The projects we do are mostly from outside freelance and include onboard 360 camera , video analytics , advertisement detection and removal, logo replacement and obfuscation.
We are sure we can deliver your project in time .
I have 12 years of SW development experience in product based MNCs in DSP Image and Video Codecs domain. Worked on POCs and Projects related to Opensource Gstreamer and ffmpeg.
Hi,
is it alright if I use python + opencv to remove the logo? Basically create some patches of the logos dynamically then match them against each frame. Also using pycuda is another option.
I have done something similar on Fedora and Arch using Python + opencv + ffmpeg although with the right codecs opencv can write the video file just fine. But yes we will need ffmpeg or v4l2 for streaming. Which streaming server are you using?