Public Lab Research note


Senior Thesis: Arduino-Raspberry Pi Root tracking system

by maya_mohanty | October 22, 2021 01:22 22 Oct 01:22 | #27954 | #27954

Hi all-

I am starting my senior thesis and will be creating an automated data collection device to track root growth. The image analysis will be performed using a raspberry pi/ camera and an arduino based system to move the raspberry pi/ camera toward the plates within an incubator. Pretty excited to start my project and will be keeping updates here on my progress!


18 Comments

Hi all, Update so far, I am preparing for a proposal for a research grant and have been researching articles for my literature review. Mechanical carousels designs that I have found are more complicated petri dish carousels however the overall concept of the automated carousel is the same for my project. These are some of the carousels that I have looked at, check them out!

http://www.letsgorobotics.com/lgr-plate-carousel-system

https://www.thermofisher.com/order/catalog/product/F00670#/F00670

Reply to this comment...


Do you want to look at commercial designs already in existence? Or would you rather stay away from them so as not to contaminate your design? Regards.

Is this a question? Click here to post it to the Questions page.

The design I am thinking about currently will be very basic/ simple due to my lack in access to expensive materials and expertise in the mechanical field however I am willing to adopt commercial designs that are currently being used!


Reply to this comment...


Hi all- Update on my project... currently thinking about the potential for disturbance off the samples using a carousel to move petri dishes toward the raspberry pi camera. The samples are plant roots in an incubator held under 100% continuous light, it is possible that moving petri dishes every 15 minutes to capture photos may introduce some disturbance. Currently brainstorming other methods of moving the plates toward the raspberry pi camera or perhaps have the camera move toward the petri dishes instead. If anyone has any ideas please let me know!

Reply to this comment...


There are some automated microbial systems. They do pretty much all the work. They incubate the samples. And instead of colony counting, they measure color, usually two or three different varieties. From that the colony equivalent can be calculated. The systems are quite expensive, and not exactly what you would want. But parts of the system would help. Sample handling and using colors to calculate colonies is going down the right path, although it isn't exactly on target. What is usually a light yellow ( for the agar) and the colonies are often white. The color of the soil is brown and the roots are usually white for the root tracking system. The root tracking system optics should be easier ( cross your fingers). And the handling of the rest of the system should also be easier ( hopefully, no sterility requirements). It's been two or three years, but let me see if the model number is in my notes.

That would be great if I could reference some parts and the way the machine is set up mechanically thank you! I may be able to make a simpler replicate using an arduino possibly. I have gone to the lab to take a look at the environment that I will be working with and these are what the plates look like and there is no background to contrast the roots unfortunately it is clear plates with no soil background.


Reply to this comment...


Hi all, heres my update- I took a trip to the plant biology lab and met with the professor that I will be collaborating with. The incubator environment that I will be working with is fridge sized and the plates are about 10 cm by 10 cm. The roots are very small and white against a clear background / see through plate. While observing the roots I have been brainstorming ways of bringing a 3-D element to the image analysis due to the fact that roots don't grow straight down in one line. The curvatures that they grow within contribute to the length of the root growth however just taking 2-D images may not fully represent the root growth in the data. My thought was creating a system in which the raspberry pi camera is able to take photos at a couple different angles slightly enough to create a stackable image that can be processed and analyzed to incorporate the 3-D element. Or maybe having two cameras vertically from each other to get the different angles rather than have the camera move vertically. Still brainstorming tho.. Here is a photo of the plate that I will be working with! image_67161345.JPG

Is this a question? Click here to post it to the Questions page.

Thanks for these great updates, @maya_mohanty! Seeing the photo of your plate made me think of the lettuce see bioassay, where stunted root growth can indicate environmental pollution. Could you see your work potentially applying to that procedure?

Is this a question? Click here to post it to the Questions page.


That is an interesting take and perspective! Potentially, yes from the environmental aspect this could be applied to measuring stress in the form of environmental pollution from the soils contaminants which has far-reaching effects on food production and could provide solutions for food hunger! Thanks for sharing!


Reply to this comment...


First, see somagen.com/products/biolumix/ The instrument was quite interesting. It used transmitted light and cylinders, not plates. My friend used it to dramatically shorten testing time for some micro testing. Don't think it would be usable for this app. This type of work requires reflected light, not transmitted. By the way, something on the page that got my attention was "no moving parts". Good luck.

Thank you, I will take a look!


Reply to this comment...


Clarification- the biolumix works on transmitted light. For this application( root tracking), you would need reflected light.

Reply to this comment...


Hi all- An update on my project, I have been researching and gathering articles related to tracking root growth and finding many different types of software that carry out this function. I have also been looking into camera sliders as well to possibly have the camera move toward the petri plates as an alternate option to the plate carousel. Here are two of the arduino powered camera sliders that I have been looking into which may be more doable design-wise.

https://makezine.com/projects/motion-control-camera-slider/

https://www.instructables.com/Make-a-Motorized-Camera-Slider/

Reply to this comment...


Hi all- My update for this week, I have been trying to work with a stepper motor and hooking it up to a raspberry pi. I might also try using the arduino with it too to test it out. My current method of learning is through the web, I have found some sites that show how to attach the device to a breadboard and the code for it but still trying to find the best way to carry out this task. The aim is to find a code that will move my camera a certain distance and stop to take a picture while continuing on its path. Once I have the code mastered, once I am able to buy parts and get my prototype started I will already have some code to work with! Looking forward to learning more about coding!

Reply to this comment...


Hi all- Update for this week, currently working on programming the stepper motor that I have from my Super Starter Kit (UNO R3 Project). I have found a youtube video along with github code that was provided and have been working on using that as practice. I have also obtained a larger stepper motor that will require research on the amount of power needed to function this motor. I have also been in search of more articles for my literature review on this topic!

https://www.youtube.com/watch?v=2wPuBVBGsCg https://github.com/hadefuwa/elegoo/blob/main/stepper_Example.ino

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


Hi all- Update for the week, I am currently trying to program a stepper motor without the use of an arduino and instead having the raspberry pi device control the motion directly. Currently working with the smaller version of the stepper motor found in my starter kit and continuing research related to plant root growth for my literature review. The links to the two guides that I am working with are below!

https://keithweaverca.medium.com/controlling-stepper-motors-using-python-with-a-raspberry-pi-b3fbd482f886

https://makersportal.com/blog/raspberry-pi-stepper-motor-control-with-nema-17

Reply to this comment...


Hi all- I have been working on programming a stepper motor for weeks and was finally able to make a stepper motor make on revolution! The issue that I was having in the beginning was the fact that it required a serial connection therefore when I ran the code nothing happened. This is the link I followed: https://www.youtube.com/watchv=2wPuBVBGsCg&ab_channel=HamedAdefuwa

Reply to this comment...


Hi all- Update for this week, I have been working on continuing programming the stepper motor. I am trying to control the stop and start function so that when I am working with the camera slider, I will be able to make the camera stop at each plate to take a picture and then move forward to the next plate to take a picture. So far I have found out that I may need to use the millis function which I have never worked with before. I found this instructable that may be helpful so I will continue playing around with this! This is the link to the instructable I am currently looking at! https://www.instructables.com/Coding-Timers-and-Delays-in-Arduino/

Reply to this comment...


Login to comment.