About Prompt
- Prompt Type – Dynamic
- Prompt Platform – ChatGPT, Grok, Deepseek, Gemini, Copilot, Midjourney, Meta AI and more
- Niche – Automated Defect Identification
- Language – English
- Category – Quality Assurance
- Prompt Title – AI Prompt for Real-Time Quality Control and Defect Detection
Prompt Details
This prompt is designed to be dynamic and adaptable for various AI platforms used in automated defect identification for quality assurance. It allows for customization based on the specific product, defect types, and imaging modalities used. The goal is to provide a robust and detailed framework for generating accurate and actionable defect detection results.
**Prompt Template:**
“`
## Real-Time Quality Control and Defect Detection Analysis
**Product:** {Product Name} (e.g., Printed Circuit Board, Woven Fabric, Metal Sheet)
**Imaging Modality:** {Imaging Type} (e.g., RGB Image, X-ray, Hyperspectral, Thermal)
**Region of Interest (ROI):** {Specify ROI if applicable; otherwise, indicate “Full Image”}
**Defect Types of Interest:** {List specific defect types} (e.g., scratches, cracks, dents, discoloration, missing components, foreign objects, weaving flaws, porosity)
**Image Data:** {Provide image data either as a file path, URL, or base64 encoded string}
**Desired Output:** {Specify desired output format} (e.g., Bounding boxes with defect labels and confidence scores, Segmentation mask highlighting defective areas, Textual description of detected defects and their severity)
**Performance Metrics:** {Specify desired performance metrics for evaluation} (e.g., Precision, Recall, F1-score, Intersection over Union (IoU))
**Acceptance Criteria:** {Define acceptable thresholds for performance metrics} (e.g., Minimum IoU of 0.8 for defect detection)
**Instructions:**
1. **Analyze the provided {Imaging Type} image data of the {Product Name} within the specified ROI.** If no ROI is provided, analyze the entire image.
2. **Identify and locate any instances of the specified defect types.** Pay close attention to subtle variations and differentiating characteristics of each defect type.
3. **For each detected defect:**
* **Classify the defect type:** Provide the specific defect label (e.g., “scratch,” “dent,” “missing component”).
* **Localize the defect:** Provide precise location information using the desired output format (e.g., bounding box coordinates, segmentation mask).
* **Quantify the defect (Optional):** If applicable, provide measurements related to the defect, such as size, area, or depth.
* **Assess the severity (Optional):** If applicable, classify the defect severity (e.g., minor, major, critical) based on predefined criteria.
* **Provide a confidence score:** Indicate the confidence level of the AI in its classification and localization of the defect.
4. **Return the results in the specified output format.** Ensure the output is structured and easily parsable for downstream processing and analysis.
5. **(Optional) If requested, provide visualizations:** Generate visualizations to highlight the detected defects on the image, such as overlaying bounding boxes or segmentation masks.
6. **(Optional) If historical data is available (e.g., previous inspection results, defect statistics):** Leverage this information to improve the accuracy and efficiency of defect detection. Specify how this data should be used (e.g., “Use this data for model fine-tuning,” “Use this data to adjust detection thresholds”).
**Example Dynamic Variables (Replace with actual values):**
* **{Product Name}:** Injection Molded Plastic Part
* **{Imaging Type}:** RGB Image
* **{ROI}:** Coordinates (x1: 100, y1: 50, x2: 200, y2: 150)
* **{Defect Types of Interest}:** Scratches, Cracks, Flash
* **{Image Data}:** /path/to/image.jpg
* **{Desired Output}:** JSON format with bounding boxes, defect labels, and confidence scores
* **{Performance Metrics}:** Precision, Recall, F1-score
* **{Acceptance Criteria}:** F1-score > 0.9
**Notes:**
* This prompt is designed to be adaptable. Modify and expand upon it based on the specific requirements of your quality control process.
* Clearly defining the defect types, desired output, and acceptance criteria is crucial for effective AI-driven defect detection.
* Experiment with different prompt variations and parameters to optimize performance for your specific application.
* When using historical data, ensure data privacy and security best practices are followed.
“`
This dynamic prompt provides a comprehensive framework for real-time quality control and defect detection across different AI platforms. By clearly specifying the product, imaging modality, defect types, and desired output, users can leverage the power of AI for accurate and efficient automated defect identification. The flexibility and customizability of the prompt ensure its applicability to a wide range of quality assurance applications.