I'm new in image/video processing, and recently I'm commissioned to run a project to measure the depth of sludge stored in a column. The sludge level varies throughout the day and there is a window on the column's wall that lets the operator to read the level through it. I plan to install a camera (possibly a GoPro), and get the real-time images (video), process them, and correlate them to the sludge level.
The whole thing looks to be easily doable, but I need some resources that help me to understand how to get the signals from the GoPro, how to process the images (videos) and then use Python or Matlab to correlate the processed images with the level.
#image_analysis #vedeo_analysis #camera #sludge_level
Dear Iman,
The best open source library for this purpose is OpenCV in phyton. There are several functions for analysis the data and correlating amongst them.
Please feel free to contact me for further discussions.
Sincerely
Behzad
I do not know how the window in your column is physically constructed. But wouldn’t it be cheaper and easier to implement, if you will measure the strength of reflected light with a simple photo sensor?
There are many analysis factors such as: PSNR, DELTA, MSE, SSIM, and VQM
Dear Iman
I would like to know firstly what type of column are you going to work on. Secondly why would you need to record real time footage from the sludge depth. You could use sludge level sensors (IFL) to measure the sludge blanket.
Amir Ehsan Torkamani Dear Amir Ehsan, the project has been defined to promote using image processing for such applications, and that is why using IFL is out of the equation (however, I still use the IFL for comparing the results). The column is clarifier tank.
Alexandre V. Baikov Dear Alexandre. the windows are made of glass, however, I don't have the details of the glass type. Also, as I mentioned to Amir Ehsan Torkamani this project has been defined to promote using image processing. The other concern with using photosensor is that it is a pointwise measurement technique, so reading the whole length of the sight glass using a single photosensor is not possible (I might be wrong on this though)
Behzad Zakeri Many thanks Behzad for the introduction, I will go through it, and get back to you if I get lost there!
Dear Iman, then I think a capture cam level system setup such as Tenevia is more suitable in your case. It possess a much more rugged build than go pro cams can work in absolutely no light environments and measures the water level using its built in software. The system is GSM, GPS, WIFI, API, FTP sharing and 4G ready and stores data on cloud in real time. The water and sludge level could be shown on the footage as a curve or level bar. Parameters such as depth, brix, color& brightness can be defined. The data requisition interval could be defined by the operator then transferred to any 3rd party software; also OS-Board proprietary software could be purchased which is able to relate various parameters to each other. Hopefully you find this useful.
Regards
PS below please find their website
https://www.tenevia.com/en/
Amir Ehsan Torkamani Many thanks Ehsan for your thorough response. I wonder if you have got any estimation on the price of this camera package (hardware+Software) as my budget is limited, and that is the reason I am leaning toward using GoPro.
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Column