tf-od-pi-test's Introduction

README

What is this repository for?

To Predict Object Detection using Tensorflow Object Detection API on Raspberry PI
Version 1.0

How do I get set up?

Install Tensorflow and other dependencies:

sudo apt-get install libblas-dev liblapack-dev python-dev libatlas-base-dev gfortran python-setuptools
sudo pip install http://ci.tensorflow.org/view/Nightly/job/nightly-pi-zero/lastSuccessfulBuild/artifact/output-artifacts/tensorflow-1.4.0-cp27-none-any.whl

Clone Tensorflow Object Detection API:

git clone https://github.com/tensorflow/models.git

Configure API using protobuf

sudo apt-get install -y protobuf-compiler
cd models/research/
protoc object_detection/protos/*.proto --python_out=.

add path to bashrc

export PYTHONPATH=$PYTHONPATH:/home/pi/models/research:/home/pi/models/research/slim

Copy trained model on pi

Train a model to run on raspberry pi using this repo

https://github.com/NanoNets/RaspberryPi-ObjectDetection-TensorFlow

Copy exported model and label file from data directory to raspberry pi

other Packages required

sudo apt-get install libjpeg-dev
sudo pip install Pillow

Predict

clone this repo and run

python ObjectDetectionPredict.py --model data/0/quantized_graph.pb --labels data/label_map.pbtxt --images /data/image1.jpg /data/image2.jpg

tf-od-pi-test's People

Contributors

Stargazers

Watchers

tf-od-pi-test's Issues

Error in print statement

python ObjectDetectionPredict.py --model data/0/quantized_graph.pb --labels data/label_map.pbtxt --images data/image1.jpg data/image2.jpg

gives the output

Traceback (most recent call last):
File "ObjectDetectionPredict.py", line 97, in
main()
File "ObjectDetectionPredict.py", line 91, in main
print("\n".join("{0:<20s}: {1:.1f}%".format(ObjectDetectionPredict_class.category_index[c]['name'], s100.) for (c, s) in zip(classes[0], scores[0])))
File "ObjectDetectionPredict.py", line 91, in
print("\n".join("{0:<20s}: {1:.1f}%".format(ObjectDetectionPredict_class.category_index[c]['name'], s100.) for (c, s) in zip(classes[0], scores[0])))
KeyError: 53.0

SSD mobile net on raspberry pi

Why I am unable to execute the detector python file in the link? it is similar to yours except the neural stick and the real-sense ....It is SSD mobile net ( written in tensorflow ) trained on coco data set.
https://github.com/kcg2015/Vehicle-Detection-and-Tracking

not found

A have a 404 for

HTTP ERROR 404
Problem accessing /view/Nightly/job/nightly-pi-zero/lastSuccessfulBuild/artifact/output-artifacts/tensorflow-1.4.0-cp27-none-any.whl. Reason:

Not Found

Recommend Projects

nanonets / tf-od-pi-test Goto Github PK