AI on the Jetson Nano LESSON 54: Recognizing and Locating Objects of Interest in OpenCV

[@krstepangovski]

Totally with you on the "cat rant" Paul - remember, that is why creators will always be few in number and highly valued - the rest are consuming BS. This stuff is hard, watching cats is easy. The few thousand watching you here are those trying to change the world, and that is who YOU are inspiring ;-) Another amazing session - thank you! Ps I have been trying to catch what darn animal keeps digging up my wife's tomato plants so am building a night-vision object detector as whatever it is, mostly comes at night. Mostly.

[@melvinsajith4448]

done the homework and got it working well ,The Objects Detection Homework

[@opalprestonshirley1700]

Fun lesson. It's great how well it does and then miss identifying an item as a refrigerator. Great rant on cats. Personally, I love dogs. Looking forward to Friday. Have a great weekend.

[@pralaymajumdar7822]

Proud of you sir from India...Grateful to you for your knowledge sharing..God will bless you surely.

[@Mircea007]

Hi Paul, This was a really fun lesson! Now it's more clear for me why you chose to go ahead with using openCV. I'm disappointed you don't receive interest you deserve, I took paid lessons for python and it's not that good compared to your lessons. And you're not only our teacher, you're our mentor!

[@paulmeistrell1726]

Fun lesson Paul. Most of my homework was done with the way the 2g lets me work. Still added blocking out screen like your cat and putting a dot in the center of detected items. Still running around 20 fps even with lots of activity and motion! Thank You..... For not liking cats... you're moving to Africa! Better be a strong box! (

[@thomascoyle3715]

Hi Paul, you forgot to go back and verify the operation of the program with the Pi Camera so I did it for you. The program correctly identified my chair, blocked it with a green box, and correctly identified my mouse and keyboard. However it thinks my ceiling fan is an airplane.

[@kamalatmeh]

lol

[@문병훈-j6r]

Comment on Cat video was touching.. Your video is much more valuable.

[@paulseidel5819]

Thanks for the tips on the Xavier paul. Proud to be one of your supporters 😁

[@rbonari]

Great lesson Paul as always !!!! We can now take actions on what we see !!!!! In the jetson nano series when you get to gpio it would be neat to have a lesson on connecting an arduino nano to the jetson nano to control stuff and read sensor into the jetson nano. You are the BEST EVER !!!!!!!

[@somebody9033]

Done homework with pretty much the same solution as you. Thank you very much for another great tutorial

[@utkarshsuthar5159]

cats part was hilarious ;)

[@jrickyramos]

Is there a way to improve the ‘cat’ dataset to create 2 different cat classes (home cats and wild cats for example) with our own images like we did for the facial recognition lesson?

[@marksholcomb]

Thanks! Great Lesson.

[@sanfinity_]

Done the homework very much like yours sir except cat part😂

[@codecage9333]

I'll send you a kitten that you can perch on the back rest of your chair and while it 'purrs' in your ear you'll pick up an additional three million viewers!

[@MrCri1tical]

I noticed when using detection from model, the box keeps flickering. Is there a way to stabilise the box especially when you have moving objects? Say could i apply opencv trackers? Would that work?

[@codecage9333]

What is the "LASER Cat Eliminator" folder on your desktop? My kittens went bonkers when they detected that folder!

[@jalopyjones6460]

Out of curiosity did you end up installing a Chicken Cam? and if so, why do i not see the live stream on Twitch?

[@majidalahmadi2133]

Please make lessons in ros robotics 🤖

[@IsaacMcC]

Hello Paul I am currently working on a Arduino project that involves a fan and a ultrasonic sound sensor. I got the code to work fine and the project works perfectly when it is plugged in to my computer, however when I have it run only on a 9v battery I can't get it to work. I have noticed when the Arduino is plugged into my pc the tx and rx pin light up but do not with the 9v. Any help? Here is the code int speedPin=5; int dir1=4; int dir2=3; int mspeed=255; const int trigPin = 9; const int echoPin = 10; long duration; int distance; void setup() { pinMode(trigPin, OUTPUT); pinMode(echoPin, INPUT); pinMode(speedPin,OUTPUT); pinMode(dir1, OUTPUT); pinMode(dir2, OUTPUT); Serial.begin(9600); } void loop() { // Clears the trigPin digitalWrite(trigPin, LOW); delayMicroseconds(2); digitalWrite(trigPin, HIGH); delayMicroseconds(10); digitalWrite(trigPin, LOW); duration = pulseIn(echoPin, HIGH); distance= duration*0.034/2; Serial.print("Distance: "); Serial.println(distance); if (distance < 35) { digitalWrite (dir1,LOW); digitalWrite (dir2,HIGH); analogWrite (speedPin,mspeed); } else { digitalWrite (dir1,LOW); digitalWrite (dir2,LOW); analogWrite (speedPin,mspeed); } }

[@Diegolike65]

i have a problem, I can´t import jetson.inference on python,

[@geeksatlarge]

Of course, your problem with cats is precisely why I like them. It's more a relationship of equality rather than subservience, LOL But yes, a show on coffee would be interesting (assuming you haven't done one already since this video was posted)

[@paulmcwhorter]

There is no equity in a relationship with the cat. The cat just hangs around to figure out what he would eat first when you die. In homes where a person dies alone, with a cat, the cat will wait a while and then start eating the owner.

[@geeksatlarge]

@@paulmcwhorter LOL

[@codecage9333]

Live eagle cams waiting for the eggs to hatch! Or maybe you should have some cats wandering around to increase viewership! :-)

[@melvinsajith4448]

cats part was hilarious ,i have 3 dogs so no cats come near my house

[@kevincespedes7757]

Is it posible to make a pan tilt camara with this object recognition ?#JetsonNano

[@TheRealFrankWizza]

Maybe 3 million people are developing cat blocking goggles, so they were testing.

[@OZtwo]

Love your videos but sorry I can not see using OpenCV since this code will automatically label and put boxes around items while not losing FPS and more moving away form the GPU to the CPU: display=jetson.utils.glDisplay(width=camera.GetWidth()+10,height=camera.GetHeight()+30) #note you as well can pass r,g,b,a to change background color of window. while display.IsOpen(): img,width,height=camera.CaptureRGBA() detections = net.Detect(img,width,height) for detect in detections: #grap data from detect display.RenderOnce(img,width,height)

[@TheRealFrankWizza]

Man, I just set up a new SD card, and now my face recognition programs run at like 1 fps. My cpu is maxed out and my gpu sits at 0%. The jetson inference stuff is working great. edit I fixed it by installing the lastest dlib.

[@codecage9333]

Missed live broadcast. Was trying to get new RasPi High Def camera working on Nano. No luck! Anyone else been successful?

[@thomascoyle3715]

Steve, there is presently no driver for the Pi HD Camera on either the Nano or the Xavier. However it works great on the Rpi 4B. I am using the 6 mm wide angle lens.

[@Cnys100]

Hi is there sombody who had some luck make this work with an own trained model? if so please coment the syntax for detectNet

[@marksholcomb]

My interest too!