SAM Quality Control is a Python-based project designed to automate and enhance quality control processes for manufacturing or service environments. The tool provides a robust framework for monitoring, analyzing, and reporting on quality metrics, ensuring that standards are consistently met.
You can visit the live version of the web application here: (SAM APP) - In maintenance
- Automated Data Collection: Seamlessly collect and process quality-related data from various sources.
- Statistical Process Control (SPC): Implement SPC charts to monitor production quality in real time.
- Defect Tracking: Track and analyze defects to identify root causes and implement corrective actions.
- Customizable Reports: Generate detailed reports tailored to specific quality control needs.
- Integration with Existing Systems: Easily integrate with existing ERP and MES systems to streamline quality control workflows.
Solving the problem of void detection on electronic components using SAM (Segment Anything Model). Given an input image or a batch of input images, our app can:
- Detect electronic components and voids
- Segment electronic components and voids
- Print a segmentation report with several metrics such as the component area and the percentage of area occupied by void.
Nb: A void is a small, empty spaces or gaps that can form within materials used in the construction of electronic component.
Here's an overview of the project structure and its components:
-
Data_Transformation: Contains the processed data required for model training and evaluation.
train/: Training dataset.test/: Testing dataset.validation/: Validation dataset.data.yaml: Configuration file for the data.
-
PCB_xray_dataset: Contains raw image data for analysis.
- YOLO Detection Model: Pre-trained YOLO models for object detection.
- SAM Segmentation Model: Pre-trained SAM models for image segmentation.
- base.html: The base HTML template with placeholders for dynamic content.
- index.html: The main entry point for the web application, extending
base.html.
- app.py: Defines the route functions for the web application.
- detection.py: Contains the script for performing YOLO object detection.
- segmentation.py: Contains the script for performing SAM image segmentation.
- report.py: Generates reports based on the detection and segmentation results.
- all.py: A script that executes all three operations (detection, segmentation, and report generation) sequentially.
-
Setup: Ensure you have all dependencies installed. You can install them using
pip:pip install -r requirements.txt