This repository contains all the necessary files and instructions to run a complete project on candy classification using digital image processing techniques.

• requirements.txt: Contains all the dependencies needed to run the project.
• cnn-model.ipynb: Notebook for building and training the classification model.
• candy-dataset.ipynb: Notebook containing all the details of the process of creating and preparing the dataset.
• Additional files: Any other essential scripts or data.

• TensorFlow: The primary framework used for model development.
• Google Colab: Platform for running Jupyter notebooks in the cloud.
• CVAT: Tool for image segmentation.
• OpenCV: For dataset creation and pre-processing.
• NumPy: For handling arrays and matrices.
• Albumentations: For data augmentation.
• Matplotlib and Seaborn: For plotting graphs.
• Scikit-learn (sklearn.metrics): For performance evaluation metrics.

1. Candy Image Dataset Creation: Collect and create your own dataset of candy images.
2. Candy Classification by Features: Organize the dataset by candy features.
3. Adjust Image Dimensions (W x H): Resize images to consistent width and height.
4. Data Annotation: Use CVAT for labeling and annotating the dataset.
5. Data Augmentation: Apply techniques to increase the variety of images in the dataset.
6. Image Filtering Functions: Implement functions to filter and preprocess images.
7. Data Normalization: Normalize image data to improve model performance.
8. Image Segmentation: Segment images to isolate regions of interest.
9. Classification: Test various methods, including CNNs and other algorithms, for candy classification.
10. Results Evaluation: Analyze the performance and accuracy of the classification models.