sarthakv7 / ai-based-indian-license-plate-detection

Indian license plate detection and character extraction using deep learning and raspberry pi.

Ver Heill ok Sæll!

Hi, I'm Sarthak Vajpayee, a self-taught Data Scientist from India who loves building apps and telling stories through data. Currently pursuing MS in Business Analytics (Data Science) from the University of Texas at Dallas. Precedingly, I worked as a Senior Data Science Consultant at , Machine Learning Engineer at Applied Roots and a Freelance Data Scientist.

I am also an open-source enthusiast and contributor. I learned a lot from the open-source community and I love how collaboration and knowledge sharing happens through open-source.

• 💼 any freelance work? do reach, email :)
• 💬 ask me about anything, i am happy to help;

languages and tools:

📊 this week i spent my time on:

Python       26 hrs 21 mins  ████████████████████░░░░░   97.77 % 
jupyter      14 mins         ████░░░░░░░░░░░░░░░░░░░░░   00.89 % 
JavaScript   12 mins         ▒░░░░░░░░░░░░░░░░░░░░░░░░   00.75 % 
Git Config   5 mins          ░░░░░░░░░░░░░░░░░░░░░░░░░   00.33 % 
XML          2 mins          ░░░░░░░░░░░░░░░░░░░░░░░░░   00.17 % 

🚧 my todoist stats:

🏆 7,982 Karma Points
🌸 Completed 4 tasks today
✅ Completed 736 tasks so far
⏳ Longest streak is 10 days

📈 my github stats

ai-based-indian-license-plate-detection's Issues

How to load the model through python script?

Hi Sarthak,

I ran the pynb file in your repo and the code seems to optimize the model, but there is no model file that is saved anywhere?
I want to use the generated model in my own python code. How to do it?

Different .csv file

I followed your tutorial until the part involving the .csv file. I downloaded what you provided here and I realize that the data are different. The one provided here has no column named 'Character' and it has only 779 rows, not 1079 like what you showed here in step 7.

Not a generalised model

The model is working well on the image used in above model,but it has failed to perform on many images downloaded from google and photo captured ones. I guess the model cannot support license plates with unequal gaps between characters and low clarity images.

csv file is not complete

image_data.csv file is not as expected. pls update the expected file

prediction not accurate

some letters are wrongly predicted example D is predicted as 0 for me as i tried same image of yours but prediction went wrong what is the reason???

IndexError: index 0 is out of bounds for axis 0 with size 0

While apply custom number plates...find_contours is not coming and error coming lik this..why

Prediction is inaccurate

Hi,
I used the same Jupyter notebook you provided with all sample data as you had specified, still, I am not getting correct predictions from the model.
I am trying AI-ML-DL for the first time so kinda new.!
See the Zip below, Please help if you can!
Untitled.zip

Error when checking target: expected dense_2 to have shape (36,) but got array with shape (27,)

Error when checking target: expected dense_2 to have shape (36,) but got array with shape (27,)

csv files also not same as in the code (Character column is not there)