Using detectNet to get coordinates of an object detected by camera

Autonomous Machines Jetson & Embedded Systems Jetson Nano
#1

Hey there, I’m working with my Jetson Nano for object detection and tracking and right now I’m just a begginer. I’ve been trying Jetson Inference’s DetectNet to see how things work. Detection algorithms are working well as I want but I couldn’t get this output as a seperate data to work with. In short I need to get these data to work with. Thanks for your help.

<detectNet.Detection object>
– ClassID: 1
– Confidence: 0.891811
– Left: 477.052
– Top: 400.135
– Right: 933.989
– Bottom: 720
– Width: 456.936
– Height: 319.865
– Area: 146158
– Center: (705.52, 560.068)

#2

Hi @myusuf2121, you can access each of the members under the detection object that are printed there - i.e. detection.Left, detection.Right, detection.Top, detection.Bottom, ect.

1 Like
(dusty-nv/jetson-inference) Accessing labels
#3

And to get the class label string, call detectNet.GetClassDesc(detection.ClassID)

2 Likes
#4

Thank you soo much, I appreciate your work by the way :) @dusty_nv

#5

@dusty_nv

Thanks! After reading your comment, I now know what I misunderstood. Doc makes sense !