File size: 4,282 Bytes
9c88d4b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 | # E-Waste PCB Component Purity & Cannibalization Index - Google Drive Package
This folder contains all project documentation, updated dataset configurations, pipeline source code, and auto-annotated YOLO label specs for the **Component Purity & Cannibalization Index** feature.
---
## 📁 Drive Folder Structure
```
google_drive_export/
├── README.md # Package documentation & usage guide
├── docs/ # Master project planning & context docs
│ ├── PROJECT_CONTEXT_AND_ACTION_PLAN.md # Master context, requirements & roadmap
│ ├── implementation_plan_component_purity_index.md # Component purity index architecture plan
│ ├── implementation_plan.md # Scraper pipeline implementation plan
│ └── scraper_plan_v2.md # Dataset pipeline architecture spec v2
├── config/ # Project configuration files
│ ├── taxonomy.yaml # Fine-grained 7-class component taxonomy
│ ├── queries.yaml # Search query matrix for e-waste & desoldered pads
│ └── config.yaml # Pipeline & aspect ratio validation settings
├── pipeline/ # Source code scripts
│ ├── auto_annotate.py # Zero-shot YOLO-World auto-annotation script
│ ├── crawler.py # URL harvesting engine
│ ├── validator.py # Image integrity & resolution validator
│ ├── processor.py # EXIF transpose & image normalizer
│ ├── dedup.py # Exact & perceptual hash deduplication engine
│ ├── manifest.py # SQLite state store abstraction layer
│ ├── queries.py # Query expansion generator
│ └── splitter.py # Cluster-aware dataset splitter
└── dataset_annotated/ # Auto-annotated dataset files
├── data.yaml # YOLO dataset configuration spec
├── auto_annotation_summary.json # Auto-annotation statistics summary
└── labels/ # 3,568 YOLO format (.txt) bounding box label files
```
---
## 🏷️ Fine-Grained Component Taxonomy (`taxonomy.yaml`)
| Class ID | Class Name | Description |
| :--- | :--- | :--- |
| `0` | `cpu_socket` | LGA, PGA, or ZIF processor socket footprints |
| `1` | `gold_finger` | Gold-plated edge connector contacts (PCIe, RAM contacts) |
| `2` | `ic_chip` | Packaged ICs and BGA microchips |
| `3` | `tantalum_capacitor` | SMD tantalum capacitors (molded yellow/black/orange) |
| `4` | `forcibly_desoldered_pad` | Exposed pads with jagged solder, blowtorch marks, torn traces, burnt flux (**fraud signal**) |
| `5` | `factory_unpopulated_pad` | Clean silkscreen outline with smooth solder dots left empty from factory (**NOT fraud**) |
| `6` | `occluded_region` | Footprints obscured by heatsinks, metal shielding, or heavy dust |
---
## 📊 Dataset Statistics (`auto_annotation_summary.json`)
* **Total Images Processed**: 4,303 images (3,923 clean normalized JPEGs)
* **Total Bounding Boxes Generated**: 3,784 draft bounding boxes
* **Class Detections**:
* `cpu_socket`: 3,570
* `tantalum_capacitor`: 121
* `ic_chip`: 67
* `occluded_region`: 10
* `gold_finger`: 7
* `factory_unpopulated_pad`: 6
* `forcibly_desoldered_pad`: 3
---
## ⚡ Next Steps & Usage Guide
1. **Upload to Google Drive**: Upload this entire `google_drive_export` folder (or `ewaste_purity_index_drive_package.zip`) to your Google Drive.
2. **Annotation Review**: Import `dataset_annotated/data.yaml` and `dataset_annotated/labels/` alongside your images into **CVAT**, **Label Studio**, or **Roboflow** for annotation review.
3. **YOLO Model Fine-Tuning**: Pass `dataset_annotated/data.yaml` to `ultralytics` to train your baseline YOLOv8 model:
```python
from ultralytics import YOLO
model = YOLO("yolov8n.pt")
model.train(data="dataset_annotated/data.yaml", epochs=50, imgsz=960)
```
|