This dataset is a meticulously curated assortment of pictures designed for the automated detection of car tire circumstances. The dataset contains pictures that seize numerous ranges of tire put on beneath totally different lighting circumstances, making it a perfect useful resource for coaching and evaluating machine studying and deep studying fashions.
- Complete Variety of Photos: 863 Stable Tires 847 Worn Tires
- Complete: 1710 Photos
- Picture Format: JPG
- Common Picture Measurement: 720×720 pixels
- Labels: Stable, Worn
The photographs have been captured beneath various lighting circumstances and from totally different angles to replicate real-world variations. Stable tires have been chosen for his or her correct tread depth and absence of deformities, representing ultimate operational circumstances. In distinction, worn tires present vital put on, equivalent to diminished tread depth, cracks, and deformities, demonstrating various levels of degradation.
This dataset serves a variety of purposes, together with however not restricted to:
- Deep Studying Initiatives:
- Coaching fashions for tire situation detection utilizing Convolutional Neural Networks (CNNs).
2. Machine Studying Research:
- Efficiency analysis with numerous machine studying fashions and algorithms.
3. Cellular Utility Improvement:
- Integration into cellular purposes for detecting the situation of car tires.
- Optimization: This dataset has been optimized to be used within the improvement and testing of fashions educated with TensorFlow and Keras. It’s also appropriate for integration with TensorFlow Lite (TFLite) for cellular purposes.
- Processing: Every picture is pre-processed to make sure uniformity, with resizing and normalization utilized as mandatory.
This dataset is freely obtainable for instructional and analysis functions. For industrial use or redistribution, please contact the dataset proprietor. When utilizing this dataset, please credit score the next reference:
Reference: Rakhmatullo Ergashev, “Automobile Tire Situation Detection Dataset,” Kaggle, 2024.
For any questions or suggestions, be happy to succeed in out through the next channels:
- E mail: [email protected]
- Kaggle Profile: rakhmatulloergashev