回复： coordinate transform--robot

[Time_Walker]

---

Hatef.fouladi 已写：

sppose you have two point  with distance of 1 cm 
you  find out that this two point distance is 3 pixcel in image 
so if you check bellow image you have this 
x step 1
y stem 0,98
unit :cm

---

hello sir:

  thank you, i understand "x step" and "Y step".

[Time_Walker]

now i know the transformation from pixel to real world unit. but i had encounter new problem in this application. how to make transformation from image coordinate to robotic coordinate? is anyone have experience on it? thank you very much~~~~

[Hatef.fouladi]

for this the easy way is that
after calibration regard 4 point in four corner of your image and also one point in center of image and steer your robot one step in y and one step in x in that points and calculate displacement of your robot in x and y axes in this case you will know what is relation between your robot moving step in x and y and and area of the environment that each pixel cover from outside
for example if you move robot one step but you find that this movement is two pixel in image so you can not order you robot to move same as one

[Time_Walker]

---

Hatef.fouladi 已写：
for this the easy way is that
after calibration regard 4 point in four corner of your image and also one point in center of image and steer your robot one step in y and one step in x in that points and calculate displacement of your robot in x and y axes in this case you will know what is relation between your robot moving step in x and y and and area of the environment that each pixel cover from outside
for example if you move robot one step but you find that this movement is two pixel in image so you can not order you robot to move same as one

---

hello master: would you like to give me a reference or example? i had check the internet and got some information. i knew the key point is the transformation between vision coordinate and robot coordinate. the key point is transformation.. thank you very much in advance.~

[Hatef.fouladi]

I can give you some link for study
http://www.diva-portal.org/smash/get/diva2:215928/fulltext01
http://www3.uji.es/~sanchiz/SplinesCalibration/
http://ftp.utcluj.ro/pub/docs/imaging/Autonomous_driving/Articole%20sortate/CThorpe/gremban_k_1988_1...

[Hatef.fouladi]

but as trick I always use for such project
after calibration I found ratio of robot step vs pixel step and then re sample image size to reach 1 ratio between robot and pixel
in this case for any pixel size you can steer your robot easly

[Time_Walker]

---

Hatef.fouladi 已写：
but as trick I always use for such project
after calibration I found ratio of robot step vs pixel step and then re sample image size to reach 1 ratio between robot and pixel
in this case for any pixel size you can steer your robot easly

---

hello master:

  ratio=robot step VS pixel step  ? or 

  ratio=robto step VS realworld step?

[Hatef.fouladi]

ratio here is means robot step VS pixel step
but because in calibration you found ratio between image pixel and realworld

so at the end you will be have  robot step vs real world 

based on image processing 
I hope I could explain it well enough to understand 

[Hatef.fouladi]

ratio here is means robot step VS pixel step
but because in calibration you found ratio between image pixel and realworld

so at the end you will be have  robot step vs real world 

based on image processing 
I hope I could explain it well enough to understand 

[Time_Walker]

---

Hatef.fouladi 已写：

ratio here is means robot step VS pixel step
but because in calibration you found ratio between image pixel and realworld

so at the end you will be have  robot step vs real world 

based on image processing 
I hope I could explain it well enough to understand 

---

hello master: i get the image coordinate and robot coordinate. and calculate the ratio between .i seem that hard to get the relationship between them. would you like to give me any suggestion?