I love your tutorials man. Thank you for all the quality content and the assumption that we’re all not pros at this already. I’m an EE student and none of this technology is taught to us so it’s up to us to teach it to ourselves if it’s the career path we want to take. These videos are great stepping stones into the this world of machine learning and AI. Once I graduate I will definitely support your patreon! Thanks again, Paul!
You deserve to be respected...No word or comment is sufficient to praise you...doing exceptionally well for the rest of the world..I am enjoying your lessons a lot lot lot in this spare time of lockdown period. Thank you..Thank you. Surely I will try to contribute as per my ability after lifting lockdown in this pandemic situation..give some time sir...please don't stop to share your knowledge.
Hi Paul, Thank you so much for all of your tutorials. The timing of finding your channel was absolutely perfect. My wife and I are old and we need to be particularly careful with COVID. My son is a Data Scientist and he got me started and you now have provided the means to really learn. We will be seeing you over at Patreon. Thanks again.
Did the home work with the cross over my face, worked super! I follow your great lesson, it was a little bit different as I did it and I learn a lot. Thank you!
Hi Paul, I have learned allot and very thankful. Managed to do the homework with no help. Because I had the facial recognition lesson done before this one, managed to lock the tracking to my face only. Really cool stuff!
Love the PI Cam. Wasn't able to get any of my 310/910/920 webcams to work (cheese app seems sees them, just blank screen and in python green screen). But what I would like is a better PI camera for darker rooms. Any suggestions?
HI Paul, still going strong, never a dull moment. Thank you for these GREAT lessons. It is a great platform to start off with. Built my own Gimbal and tracking is amazing what you can do. Question - I want to flip out the PiCam and Logitech with Sony Alpha5100 or GoPRo and very curious to test it with the Flir View Pro - these are all cameras I am currently using. I've installed the Arducam CSI to HDMI Cable Extension Module connecting the Cameras - I know I was pushing my luck here - Any tips on this one
Hi, Paul. I wanted to start learning Python with your tutorials but I was unable to find the link for the serial port extension. It looks like they've been moved to a new address. Do you know the new link? Much appreciated if you have the time to help me out here. Many thanks.
Hello Paul, Great lesson as always !!!! Very outstanding !!!!!! I do have one question. Instead of errorPan=Xcent-dispW/2 why wouldn't it be errorPan=dispW/2-Xcent. Also it seems errorTilt=Ycent-dispH/2 should be errorTilt=dispH/2-Ycent from the way you explained the process? Thanks again for all your great lessons !!!!!!!!
It just depends on your frame of reference. You could do as you suggest, but then the following if statements would need to be done differently. Just comes down to in your mind what is a positive vs negative error.
Itsme, we are forced to upgrade, but you need to do it with me in the next few lessons. We are forced to because openCV face_recognition library I can not get installed on our jetpack. Hence, since we want to IDENTIFY faces, we must have the face_recogntion library which forces the upgrade. So, we will spend a few lessons getting configured for the new jetpack. All our old code should work, except the cv2.findCountours() command returns two parameters, so when we used that line of code, it has to be edited to remove the first _, Everything else works as expected, but we have to go in and install code-oss and all that type of stuff. But, then things get really fun.
Hey Paul, was wondering why my camera moved in an inverted diagonal motion? I’m using the correct code as far as I can tell and I still can’t figure it out
I believe it has something to do with the flip=2, I had to change it to flip=0 because my pi camera is mounted upside down. Is there any way to invert the tilt so when I flip my screen to the correct orientation it doesn’t mess with the servo data
Ah I got left in the dust after lesson 11 it was fun though but I had to move on to another card and ros melodic but I might use something similar in ros eventually .👍😁 good video.
Once the pan or tilt error was determined to be>15, I then had to add a further test to check the sign on the error in order determine whether to add or subtract 1/50th of the pan error to correct the error. I'm sure the sign is important!
I dont think that if statement is needed, depending on how you define error. If error = (desired - actual), then if actual is bigger than desired, error will be negative, and just add a fraction of the error. If actual is smaller than error, then error is positive and just add a fraction of the error. In either case, you add a fraction of the error, and the sign of error itself takes care of iteself, if that makes sense
Excellent series Paul. Learning python and opencv very quickly thanks to you. One problem I've encountered though is I have tracking working on the Picam but on the webcam I have it won't work. I am using a ELP-USBFHD01M-SFV (5-50mm) webcam and keep getting:- cap_gstreamer error.cpp(933), can anyone help with this? I am not that versed in programming yet (crusty old mechanical guy) so plain english would be good. Cheers, once again kudos for this series. Forgot to mention using jetpack 4.3 and cv2 - 4.1.1.
try having just the USB connected, and run 'cheese'. This is a simple program to show web cam. If cheese does not find the WEB cam, then it might not be compatible.
Also, WHEN USING THE WEBCAM, THE LINE cam=cv2.VideoCapture(1) MUST BE REPLACED WITH cam=cv2.VideoCapture(’dev/video1’) FOR JETPACK VERSIONS 4.3 AND BEYOND.
I love your tutorials man. Thank you for all the quality content and the assumption that we’re all not pros at this already. I’m an EE student and none of this technology is taught to us so it’s up to us to teach it to ourselves if it’s the career path we want to take. These videos are great stepping stones into the this world of machine learning and AI. Once I graduate I will definitely support your patreon! Thanks again, Paul!
You deserve to be respected...No word or comment is sufficient to praise you...doing exceptionally well for the rest of the world..I am enjoying your lessons a lot lot lot in this spare time of lockdown period. Thank you..Thank you.
Surely I will try to contribute as per my ability after lifting lockdown in this pandemic situation..give some time sir...please don't stop to share your knowledge.
Hi Paul, Thank you so much for all of your tutorials. The timing of finding your channel was absolutely perfect. My wife and I are old and we need to be particularly careful with COVID. My son is a Data Scientist and he got me started and you now have provided the means to really learn. We will be seeing you over at Patreon. Thanks again.
Thank you for the kind words!
Did the home work with the cross over my face, worked super!
I follow your great lesson, it was a little bit different as I did it and I learn a lot. Thank you!
No problem with long lesson with results like this. Always a great time. Thanks Paul.
double chest bump! Again thanks for all the great lessons!
Hello Paul, thanks for the effort in the making of your Jetson Nano Videos! Great Job!
Hi Paul,
I have learned allot and very thankful. Managed to do the homework with no help. Because I had the facial recognition lesson done before this one, managed to lock the tracking to my face only. Really cool stuff!
sorry, skipped the eye video and did my homework even before knowing was going to need it soon. :) Great videos fyi.
This is pretty cool. Don’t ever stop Paul!
Great project. Really enjoying working my through these. Thanks Paul
great lesson, can't wait until my jetison arrives to go all lessons again
Count down begins for Super Excitement !! Keep Going Paul !!
Will do!
back into it again, got my servos and loving it, thanks Paul
This is another great lesson I ever wanted. Very appreciate it.
woooowwww much awaited, Paul now these lessons are getting lot more excited. Keep going 👍
A little late but happy to be joining the party! Thanks Paul!
Welcome. Can not remember, do you have your jetson nano gear yet?
Paul McWhorter nope. It’s still on its way but I thought it’d be good to watch the lessons anyway. 👍
Hey, Paul! Great videos!! That mysterious moveWindow problem: you had nanoCam spelled namoCam.
Love the PI Cam. Wasn't able to get any of my 310/910/920 webcams to work (cheese app seems sees them, just blank screen and in python green screen). But what I would like is a better PI camera for darker rooms. Any suggestions?
Great lesson Paul, as always - keep it up!
HI Paul, still going strong, never a dull moment. Thank you for these GREAT lessons. It is a great platform to start off with.
Built my own Gimbal and tracking is amazing what you can do.
Question - I want to flip out the PiCam and Logitech with Sony Alpha5100 or GoPRo and very curious to test it with the Flir View Pro - these are all cameras I am currently using.
I've installed the Arducam CSI to HDMI Cable Extension Module connecting the Cameras - I know I was pushing my luck here - Any tips on this one
Great video sir, but can I use the jetson nano with the AX-12 servos?
i have something understand that when you set the abs（values） for offset， how can the pan or the tilt know to change the left or right、top or down？？
Another great lesson. Thank you
My pleasure!
Awesome Lesson, great to play with jetson : )
Glad you like it!
Note: when testing be sure to close windows. Camera saw someone outside and turned camera 180 watch them mowing lawn. :)
Hi, Paul. I wanted to start learning Python with your tutorials but I was unable to find the link for the serial port extension. It looks like they've been moved to a new address. Do you know the new link? Much appreciated if you have the time to help me out here. Many thanks.
Exciting !!!!!
My webcam has gotten all of a sudden super low fps. 1-2 or something like that.
Any idea of what can have caused it?
Hello Paul,
Great lesson as always !!!! Very outstanding !!!!!! I do have one question. Instead of errorPan=Xcent-dispW/2 why wouldn't it be errorPan=dispW/2-Xcent. Also it seems errorTilt=Ycent-dispH/2 should be errorTilt=dispH/2-Ycent from the way you explained the process? Thanks again for all your great lessons !!!!!!!!
It just depends on your frame of reference. You could do as you suggest, but then the following if statements would need to be done differently. Just comes down to in your mind what is a positive vs negative error.
Thanks much for your response !!!!!!
In the live chat you mentioned a few negative things about 4.3. Is it really worth upgrading for the rest of this course?
Itsme, we are forced to upgrade, but you need to do it with me in the next few lessons. We are forced to because openCV face_recognition library I can not get installed on our jetpack. Hence, since we want to IDENTIFY faces, we must have the face_recogntion library which forces the upgrade. So, we will spend a few lessons getting configured for the new jetpack. All our old code should work, except the cv2.findCountours() command returns two parameters, so when we used that line of code, it has to be edited to remove the first _, Everything else works as expected, but we have to go in and install code-oss and all that type of stuff. But, then things get really fun.
Hey Paul, was wondering why my camera moved in an inverted diagonal motion? I’m using the correct code as far as I can tell and I still can’t figure it out
I believe it has something to do with the flip=2, I had to change it to flip=0 because my pi camera is mounted upside down. Is there any way to invert the tilt so when I flip my screen to the correct orientation it doesn’t mess with the servo data
How can we use it on embedded vision. Is it able to upload it to Arduino.... Etc
Ah I got left in the dust after lesson 11 it was fun though but I had to move on to another card and ros melodic but I might use something similar in ros eventually .👍😁 good video.
Once the pan or tilt error was determined to be>15, I then had to add a further test to check the sign on the error in order determine whether to add or subtract 1/50th of the pan error to correct the error. I'm sure the sign is important!
I dont think that if statement is needed, depending on how you define error. If error = (desired - actual), then if actual is bigger than desired, error will be negative, and just add a fraction of the error. If actual is smaller than error, then error is positive and just add a fraction of the error. In either case, you add a fraction of the error, and the sign of error itself takes care of iteself, if that makes sense
I finally got around to doing this with face ID. It's actually a lot easier than this.
Sir I done several projects using raspberry pi so should I update to jetson nano
YES! The nano is a true AI machine. The pi will run out of steam.
Excellent series Paul. Learning python and opencv very quickly thanks to you. One problem I've encountered though is I have tracking working on the Picam but on the webcam I have it won't work. I am using a ELP-USBFHD01M-SFV (5-50mm) webcam and keep getting:- cap_gstreamer error.cpp(933), can anyone help with this? I am not that versed in programming yet (crusty old mechanical guy) so plain english would be good. Cheers, once again kudos for this series. Forgot to mention using jetpack 4.3 and cv2 - 4.1.1.
try having just the USB connected, and run 'cheese'. This is a simple program to show web cam. If cheese does not find the WEB cam, then it might not be compatible.
Also, WHEN USING THE WEBCAM, THE LINE
cam=cv2.VideoCapture(1)
MUST BE REPLACED WITH
cam=cv2.VideoCapture(’dev/video1’)
FOR JETPACK VERSIONS 4.3 AND BEYOND.
can not wait to get into deep learning as OpenCV simply is taking up way to much processor power here at 8-16 FPS (20FPS at 320x240)
In 9.36, your window name was ('nanoCam' ) but in moveWindow you wrote it as (' namoCam') as namo not nano. That's why it didn't move . ;)
Great Lesson Paul. The crashes are the code not your cosmetics (
Whats your email my dad would like to send you an email about the arduino.