Tagging the images which have the road signs
Input | Output |
---|---|
{ 'tag': 'road_sign', 'score': 0.9429549 } |
|
{ 'tag': 'non_road_sign', 'score': 0.7540367 } |
to install the road sign tagging model, you need Python 3.7.7
git clone https://github.com/liang6261515/road_sign_tagging.git
cd road_sign_tagging
pip3 install -r requirements.txt
wget https://github.com/fchollet/deep-learning-models/releases/download/v0.4/xception_weights_tf_dim_ordering_tf_kernels_notop.h5
download the pretrain model of cheque detection from this url https://drive.google.com/file/d/1lGkgRc1VU1tVJPiMUSVyCV2Elm1Gy8_-/view?usp=sharing to the folder road_sign_tagging
down load an image of road sign
wget https://c8.alamy.com/comp/FPMFN8/road-signs-dubai-uae-FPMFN8.jpg
this image looks like
import the model in python3
from road_sign_tagging import road_sign_tagging
run the tagging program
road_sign_tagging('road-signs-dubai-uae-FPMFN8.jpg')
you will see the following output:
{'tag': 'road_sign', 'score': 0.9429549}
now lets download another example of image which does not have any road sign
wget https://live.staticflickr.com/4169/34649653051_9e31056a0a_b.jpg
it looks like
run the tagging program
road_sign_tagging('34649653051_9e31056a0a_b.jpg')
and it has the following output since it has no road sign
{'tag': 'non_road_sign', 'score': 0.7540367}
performing OCR to extract the road and location names from the road signs and map them to the goe-location, so that the road signs can searched on a map
feel free to contact me if you have any problem with this package or you are hiring data scientist/AI engineer. I am actively looking for data science/AI related jobs
My email: yanliang2345@outlook.com