Computer Vision IOT Set up

X

XDabxdoubx

Active Member
I am building a set up for computer vision research that 3D models your plants for you in real time based on the plants strains and genetics and essentially forecast it. It's a work in progress and I plan to make it open source sense I dont have the money to start my own business or live in a legal state to do this. A couple raspberry pis hooked up to some dwc drip buckets and some usb camera for photogrametry. I have a patent for it given to me by the N.S.F but I just turned 20 and didnt have the cash to even pay for it to be official till now.
 
X

XDabxdoubx

Active Member
I may combine this with my autonomous ph meter to add for further automation and data collection. If anyone is interested in some source code or how to implement it on a larger scale. Message me, I have a degree in sustainable agriculture and would love to work on this outside of ohio.....ya know..iron curtains
 
Renfro

Renfro

Well-Known Member
XDabxdoubx said:
I am building a set up for computer vision research that 3D models your plants for you in real time based on the plants strains and genetics and essentially forecast it.
Click to expand...
Ok, I am curious how your algorithm will account for the many variables involved. For example, environmental variables, feed variables, lighting variables, even things like pests and all the other variables. It seems a mighty big mountain to climb. I am a little confused on the modeling part.

Now you mention automation, thats something many people would really like. For example I would love something I could have say 50 pH probes, the kind that you stick in soil, all taking real time readings on all my plants, logging that data. Perhaps even reacting and dosing nutrients to an adjusted pH. Is something like that doable?
 
ChiefRunningPhist

ChiefRunningPhist

Well-Known Member
Renfro said:
Ok, I am curious how your algorithm will account for the many variables involved. For example, environmental variables, feed variables, lighting variables, even things like pests and all the other variables. It seems a mighty big mountain to climb. I am a little confused on the modeling part.

Now you mention automation, thats something many people would really like. For example I would love something I could have say 50 pH probes, the kind that you stick in soil, all taking real time readings on all my plants, logging that data. Perhaps even reacting and dosing nutrients to an adjusted pH. Is something like that doable?
Click to expand...

This might be what you're after...
 
X

XDabxdoubx

Active Member
Renfro said:
Yeah but I need individual soil probes, most important part is documenting individual pH in the soils. For me anyway
Click to expand...
I dont know about 6 different individual soil ph probs but I have made an autonomous ph sensor for a single hydroponics reservoir for a shared dwc system. The 3d modeling is done with photogrametry similar to how drones 3d model. The major task is data collection and gathering the computer vision data. I have built a few systems for other people to help pay for college but now it's my turn to build my own so I'm going full flex on my budget.lighting is measured with a inferred sonar raspberry pi camera that can detect light strength and temperature and will 3d model that in the app for you in an AR type fashion. I'm not totally sure this is been done before but I'm getting everything together.
 
X

XDabxdoubx

Active Member
I know what I want is ambitious but it would be good for the community to have access to this open source information. I am a newbie and as such will take any sage advice given to me.
 
2

2com

Well-Known Member
Renfro said:
Yeah but I need individual soil probes, most important part is documenting individual pH in the soils. For me anyway
Click to expand...
Individual soil probes can be had. I think I shared a link to one (for arduino builds) just as a "hey, these exist" post. What about pressure load cells? Could someone rig those underneath planter pots act as scales and either alert or dose when the weight falls under a certain threshold. You'd have to dial in "dry" and "wet" weights and stuff, but you'd have to do that at least to a degree with moisture sensors too, I think.

There's a guy who's using digital scales (the kind they use at commercial couriers) under his planters to let him know when they're dry.
 
Dryxi

Dryxi

Well-Known Member
Renfro said:
Yeah but I need individual soil probes, most important part is documenting individual pH in the soils. For me anyway
Click to expand...
Testing the soil water pH would be a hard job for a sensor I think. Pretty sure those sensors need a more liquid solution to pH rather than a mostly solid (soil) with only a little liquid. Best you could test for there is the moisture level in the soil, I think. I really don't know enough about the sensors you could use.

Possible hack for that could be to put the pots on a stand, like 1"x1" with something to catch the runoff, you could have it auto water a certain amount when the moisture level reaches a certain number, then have a probe in the basin of runoff to test that pH. I don't know if you do the whole water to runoff though. But that would only work if you also had an automated method to the remove the water from the basin so it can test it again at the next watering.
 
Last edited:
Renfro

Renfro

Well-Known Member
Dryxi said:
Testing the soil water pH would be a hard job for a sensor I think. Pretty sure those sensors need a more liquid solution to pH rather than a mostly solid (soil) with only a little liquid. Best you could test for there is the moisture level in the soil, I think. I really don't know enough about the sensors you could use.

Possible hack for that could be to put the pots on a stand, like 1"x1" with something to catch the runoff, you could have it auto water a certain amount when the moisture level reaches a certain number, then have a probe in the basin of runoff to test that pH. I don't know if you do the whole water to runoff though. But that would only work if you also had an automated method to the remove the water from the basin so it can test it again at the next watering.
Click to expand...
I know that Apera, Bluelab and Hanna all make pH meters for soil, i have them all lol. They work great. Apera's higher end soil pH meter uses a probe that is separate from the unit and has a BNC connector so something like that would work but they aren't cheap.
 
X

XDabxdoubx

Active Member
I guess the end goal is to get relatively affordable parts and sensors together so we can have an opensource next level environment on our hands. Craft cannabis should be able to use the same technology to aid the craft that big cannabis does. I'm really busy at work atm so this projects on my back burner but I want to create a github so I can upload some cad files
 
Roger A. Shrubber

Roger A. Shrubber

Well-Known Member
Renfro said:
Ok, I am curious how your algorithm will account for the many variables involved. For example, environmental variables, feed variables, lighting variables, even things like pests and all the other variables. It seems a mighty big mountain to climb. I am a little confused on the modeling part.

Now you mention automation, thats something many people would really like. For example I would love something I could have say 50 pH probes, the kind that you stick in soil, all taking real time readings on all my plants, logging that data. Perhaps even reacting and dosing nutrients to an adjusted pH. Is something like that doable?
Click to expand...
you forgot the biggest variable...how smart or stupid the grower is...i'd really like to see the algorithm that accounts for that
 
You must log in or register to reply here.
Top