Training Dataset: This is the dataset used to train the model. The training dataset includes images that have been accurately labeled with the location and type of road damage, such as cracks, potholes, and depressions. The data has been carefully prepared to ensure that all types of damage are correctly classified. Each image is accompanied by a label indicating the type of damage and its location on the road surface. The training dataset is diverse in various scenarios, including different lighting conditions, road types, and varying levels of damage severity. Having sufficient diversity in the training dataset helps the model learn the different characteristics of road damage in real-world environments.
Testing Dataset: The testing dataset is an independent dataset that is not used during the model training process. This data is used to evaluate the model's performance after the training process is complete. The purpose of the testing dataset is to assess the model's generalization ability, which means its ability to accurately identify and classify road damage it has not encountered during training.
The testing dataset is selected with real-world scenarios and conditions that the model will need to handle in the deployment environment, such as changing weather, low light, or rare types of damage.
Dataset division: The dataset uses an 80% training set and a 20% testing set, ensuring that the data in both sets is diverse and sufficiently balanced.
4.3.2. Set training parameters
- Network architecture: Choose a suitable network architecture, in this case, you can use YOLOv11n-seg or variants of YOLOv11n depending on the specific requirements.
- Learning Rate: This is an important parameter in the training process, controlling the learning rate of the model. It is often necessary to set an initial learning rate and
can use learning rate tuning techniques such as decay, cyclic learning rate to improve training performance.
- Batch Size: The number of data samples used in each update of the network's weights. The larger the batch size, the faster the training process, but also needs to consider computational resources. - Epochs: An epoch is a pass through the entire training data set. The number of iterations depends on the data size and the complexity of the model.
Figure 4.5: Model Training
-Optimizer: There are many types of optimizers such as Adam (Adaptive Moment Estimation), SGD (Stochastic Gradient Descent), and RMSProp (Root Mean Square Propagation). Choosing the right optimizer helps to increase the speed and efficiency of the training process.
- Loss Function: The loss function needs to be chosen to accurately reflect the level of deviation between the model's prediction and the actual label. For object recognition and segmentation problems, functions such as Mean Squared Error (MSE) or Intersection over Union (IoU) loss are often used.
- Regularization: To avoid overfitting, regularization techniques such as L1/L2 regularization or dropout can be used.
- Early stopping: Use the early stopping technique to stop the training process when there is no significant improvement on the validation dataset.
- Testing and tuning: After setting the parameters, it is necessary to test and tune to optimize the performance of the model before official training.
The training parameter setting process requires careful control and tuning to ensure the model achieves the best performance and avoids overfitting.
4.3.3 Processing results and reporting Processing results from the model
Once the pavement damage recognition and segmentation model has been implemented, the results are processed to generate the necessary decisions or reports.
- Result segmentation: Based on the bounding boxes or masks generated by the model, determine the locations and types of damage on the pavement.
- Result filtering: Thresholds can be applied to eliminate incorrect or unnecessary detections. For example, removing small or unimportant objects.
- Feature calculation: If necessary, additional features can be calculated from the recognition and segmentation results. For example, the depth of road rut, the length of cracks, etc.
- Image tagging: Re-tag the images with information about the locations and types of damage that have been recognized for easy review and assessment later.
Figure 4.6: Google Colab Image Test Results
Figure 4.7: Google Colab Video Test Results
Generate reports and send notifications
- Generate reports: Based on the processing results from the model, automatically generate reports on the condition of the pavement, including location, type of damage, and severity.
- Send notifications: The system can automatically send notifications or warnings to managers or pavement maintenance teams. Notifications can include content about detected problems that need to be resolved.
- Update management system: Results and reports from the model can be updated directly into the road management system for monitoring and planning repairs and maintenance.
- Overall assessment: Synthesize results from the recognition and segmentation model to evaluate the overall condition of the pavement, make decisions about repairs, maintenance or improve the management system.
- Data storage: Results, reports and notifications need to be stored for later retrieval and analysis.
The process of processing results and generating reports from pavement damage identification and segmentation models helps automate the process of managing, monitoring and maintaining roads, while providing complete and accurate information to managers and maintenance teams.