I am newbie in robotics but familiar with computer programming. I want to develop system for a real world application related with agriculture. Machine should be able to harvest fruits/flowers from plants. For machine vision I want to use openMV camera , python lang as software and arduino processor so that I can calculate exact coordinates of target. Broadly steps should include 1) Install camera in PC 2) Develop algorithm for coordinates of flowers/fruits 3) Load compiled code to arduino. After which robot design will follow. Does this is correct approach or I am missing something ?
Sure, what’s the Arduino for however? The OpenMV Cam replaces the need for that to be the main processor.
Thanx kwagyeman. Does OpenMV will work in following real world agriculture farm - 1) OpenMV should be able to locate target from a distance of 3 to 4 feet length 2) Plants have height up to 6 feet 3) Targeted fruits/flowers may be in overlapped 3) Bright sunny day light. If OpenMV is the not solution then what would be choice ? 3D machine vision ?. Everything now is at scratch level
We can easily do the color tracking for you. I’m just noting that the camera doesn’t need an Arduino in the loop for control. If you plan to use an Arduino the Arduino should be a slave processor for motor control or something like that.
With the discussion here , I am able to search precisely on net. One more stupid question , after locating target within image which is received from opemmv, how end effector reaches at exact targeted area without human intevention? I mean what programming logic/tools I should use for it ?
Please read up on inverse kinematics.
Basically, you can derive equations for each arm joint that take a point in 3d space such that when those equations are feed a point they move the arm to it. The math is a little tricky though… You may be able to find some software packages that spit out the equations for you.