search

google-coral / pycoral

Python API for ML inferencing and transfer-learning on Coral devices
https://coral.ai
Apache License 2.0
347 stars 144 forks source link

New, looking for way to detect and then classify in a simple file.... #70

Open Swiftnesses opened 2 years ago

Swiftnesses commented 2 years ago

Description

Hello,

I've just downloaded pycoral and am really enjoying learning this new tool.

Essentially, I'm trying to combine the detect and classify examples together:

Look at an image (webcam jpg capture 1920x1080) and firstly identify if there are 'birds' present. If there are, I'd like to somehow utilise the bbox coordinates and run them though MobileNet V2 to classify the bird type.

I'd be really grateful if someone could get me started, I've researched for hours how to combine the functionality without luck...

Thank you.

Click to expand! ### Issue Type Support ### Operating System Linux ### Coral Device M.2 Accelerator A+E ### Other Devices _No response_ ### Programming Language Python 3.9 ### Relevant Log Output _No response_
Namburger commented 2 years ago

@Swiftnesses maybe you can take a look at manufacturing demo? https://github.com/google-coral/demo-manufacturing

The right side does a detection on the box and then run it through and classifier to see if the image in the bbox is a good or bad apple :)

Swiftnesses commented 2 years ago

Hi @Namburger - I'll take a look but as a beginner, my first impression is that I'll not stand a chance understanding it!

I'm currently trying to use the box coordinates to resize the image, frustrating... :(