How to run tesseract ocr on inference engine

Autonomous Machines Jetson & Embedded Systems Jetson Nano
1

Hardware used : jetson nano 2gb
Programming lang: python on VS code

We are running tesseract- OCR version 4. on live video stream.The output runs at less than 1 fps. We want to run tesseract on inference engine to increase the speed. Is this possible? if yes how?

3

Hi,

It depends on the layer used in the tesseract’s model.
You can find the TensorRT support matrix in the following document:

And here is the performance benchmark for Jetson.
You can check if the speed-up can meet your requirement or not first:

Thanks.