PCB Defect Classification run on Jetson Orin NX using deepstream without any output

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Please provide the following information when requesting support.

• Hardware: Jetson Orin NX 16G

IMG_4600
IMG_46001007×302 62.7 KB

• Process to reproduce:
download PCB defect classification models from PCB Defect Classification | NVIDIA NGC
following the guide to install deepstream and run:
ds-tao-classifier -c pgie_multi_task_tao_config.txt -i file:///home/nvidia/pcb682.jpg
sample config pgie_multi_task_tao_config.txt (1.7 KB)

test image pcb682
test image pcb682684×1064 43.7 KB

result:
no any detection outputs in output image or command line
op17.txt (3.0 KB)

Is this correct or any thing wrong?

3

Have you set the correct label file? “/home/nvidia/deepstream_tao_apps/configs/multi_task_tao/label.txt” is not for this model. Please read the PCB Defect Classification | NVIDIA NGC introduction.

4

yes, we create this label according to the introduction

5

clarified, this model could only detect images which focused on one single material. Full PCB may not fit for it.

6

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.