Jetson inference car detect

yujy9014 · July 29, 2020, 5:54am

We are conducting vehicle detection by installing the Jetson reference of Jetson Nano. I’d like to designate an area so that vehicles can be detected only in certain areas, is there a good way?

dusty_nv · July 29, 2020, 5:37pm

Hi yujy9014, one such way is by cropping the input video frame to your Region of Interest (ROI), and sending that cropped frame to detectNet. Otherwise you could just filter/reject the detected objects from the full frame that lie outside of your ROI.

See here to use the cudaCrop() function to crop the image:

Topic		Replies	Views
Detect.net for determined part of video Jetson TX2 jetson-inference	4	452	October 18, 2021
How can I only detect objects that are located in the center region of the screen? Jetson Nano jetson-inference	7	559	January 19, 2022
Detectnet.py region detection Jetson Xavier NX jetson-inference	2	646	January 19, 2022
ROS Deep Learning cropping image Jetson Nano cuda , jetson-inference	4	531	October 15, 2021
DetectNet: Crop image with detected object(bbox) and run a second detection on it Jetson Nano jetson-inference	3	738	September 13, 2021
Using Jetson Nano to produce an Output after conducting Object Detection Deep Learning (Training & Inference)	0	409	October 13, 2020
Object detection - location on image Jetson Nano jetson-inference	2	675	January 26, 2022
Less info please! I need my network to show only one bonding box at the time Jetson Nano jetson-inference	3	458	October 15, 2021
Bounding box ssd jetson nano Jetson Nano jetson-inference	3	1303	October 15, 2021
Is there a way to detect and track objects within multiple ROIs of the frame? Jetson Orin Nano jetson-inference	2	233	January 23, 2024

Jetson inference car detect

Related topics