Presented By :-
Divyanshi Sharma (02696302819)
Ayushi Choudhary (00396307320)
Suraj Suman (01496307320)
Traffic Sign Recognition using
CNN
Objective:-
• Creation of a simple running program in Python that helps detect,
classify and recognize a traffic sign from a live input.
• The program will match the image over different filters using a CNN
Model to give an accurate output.
Methodology
1.Collect a comprehensive dataset of traffic
sign images.
2. Preprocess the dataset by resizing,
normalizing, and augmenting the images.
3. Split the dataset into training, validation,
and testing sets.
4. Design the architecture of the CNN.
5. Compile the CNN model with the
appropriate loss function and optimizer.
6. the CNN using the training dataset.
7. Tune the hyperparameters of the model
using the validation set.
8. Evaluate the performance of the trained
9. Deploy the model for practical use.
The main process behind this is using CNN for proper recognition of signs , as it gives the accurate output , all
while working it across various layers and filters.
Here’s an example of how CNN works for better understanding:-
Outcome:
• This project will create a CNN model to identify traffic signs and will
also classify them with approx 95% accuracy.
• we will observe the accuracy and loss changes over a large dataset.
Application :
• used in Automatic Driving vehicle as a driving assistance.
• used in map service for effective, fast and accurate traffic data updates.
• Used in Automated Industry for accuracy and assistance.
Challenges :
• vast variety of traffic sign
• Susceptibiity to changes under Natural condition
• The whole process of detection and update must be fast and accurate.
• Huge complexity of the forms of traffic sign.

Traffic_Sign_Recognition_Using_CNN_-_PPT.pptx

  • 1.
    Presented By :- DivyanshiSharma (02696302819) Ayushi Choudhary (00396307320) Suraj Suman (01496307320) Traffic Sign Recognition using CNN
  • 2.
    Objective:- • Creation ofa simple running program in Python that helps detect, classify and recognize a traffic sign from a live input. • The program will match the image over different filters using a CNN Model to give an accurate output.
  • 3.
    Methodology 1.Collect a comprehensivedataset of traffic sign images. 2. Preprocess the dataset by resizing, normalizing, and augmenting the images. 3. Split the dataset into training, validation, and testing sets. 4. Design the architecture of the CNN. 5. Compile the CNN model with the appropriate loss function and optimizer. 6. the CNN using the training dataset. 7. Tune the hyperparameters of the model using the validation set. 8. Evaluate the performance of the trained 9. Deploy the model for practical use.
  • 4.
    The main processbehind this is using CNN for proper recognition of signs , as it gives the accurate output , all while working it across various layers and filters. Here’s an example of how CNN works for better understanding:-
  • 5.
    Outcome: • This projectwill create a CNN model to identify traffic signs and will also classify them with approx 95% accuracy. • we will observe the accuracy and loss changes over a large dataset.
  • 6.
    Application : • usedin Automatic Driving vehicle as a driving assistance. • used in map service for effective, fast and accurate traffic data updates. • Used in Automated Industry for accuracy and assistance.
  • 7.
    Challenges : • vastvariety of traffic sign • Susceptibiity to changes under Natural condition • The whole process of detection and update must be fast and accurate. • Huge complexity of the forms of traffic sign.