AI & Computer Vision for Classification
Aim
To accurately identify and classify objects within images using computer vision and AI techniques, enabling the achievement of project deliverables through the evaluation and comparison of multiple solution approaches.
Solution B will look to take the following approach inspired by PatchX:
Use the given datasets
Multiple coins per image
Classify coin variations from dataset 1
Train a model using these classifications
Use this trained model to achieve deliverables on dataset 2
Solution A will look to do the following using some similar methods from ReWeave:
Create new dataset for each coin variation
One coin per image
Soley use this smaller database for training and validation
Create additional dataset for testing our trained model
Use cv2 features to isolate coins
Individually identify each coin and classify, then achieve the deliverables.
