azure-ai-vision-imageanalysis-java
Build image analysis apps with Azure AI Vision Java SDK
Need to add image captioning, OCR text extraction, object detection, or smart cropping to your Java application. This skill provides ready-to-use code patterns for integrating Azure AI Vision SDK with proper credential management and error handling.
Download the skill ZIP
Upload in Claude
Go to Settings → Capabilities → Skills → Upload skill
Toggle on and start using
Test it
Using "azure-ai-vision-imageanalysis-java". Generate caption for image.jpg
Expected outcome:
The Azure AI Vision SDK generates a caption like: 'A group of people sitting at a conference table discussing a presentation' with a confidence score of 0.92. The code shows how to extract both the text and confidence value.
Using "azure-ai-vision-imageanalysis-java". Extract text from receipt.png
Expected outcome:
OCR results include structured text blocks with line and word-level details. Each word shows its text, confidence score (e.g., 0.98), and bounding polygon coordinates for precise positioning.
Using "azure-ai-vision-imageanalysis-java". Detect objects in photo.jpg
Expected outcome:
Object detection returns an array of detected objects. Each object includes tags with names (e.g., 'person', 'laptop') and confidence scores, plus bounding box coordinates (x, y, width, height) for drawing detection boxes.
Security Audit
SafeThis is a legitimate Azure AI Vision SDK documentation skill for Java. Static analysis flagged many patterns but all are false positives: (1) External commands detected at lines 15-268 are Maven XML and Java code examples, not shell commands. (2) Network URLs at lines 102, 269, 281 are example/demo URLs and documentation links. (3) Environment variable access at lines 32-33 uses System.getenv() for secure credential management - this is the recommended secure practice. (4) High/blocker findings for 'weak cryptographic algorithm' are YAML metadata and table formatting, not cryptographic code. The skill uses proper security practices by reading Azure credentials from environment variables rather than hardcoding them.
Low Risk Issues (4)
Detected Patterns
Quality Score
What You Can Build
Automate document digitization
Extract text from scanned documents, receipts, and invoices to create searchable digital archives. Use OCR to convert paper documents into editable text.
Enhance e-commerce product catalogs
Automatically generate product image captions and detect objects for better search indexing. Create consistent smart-cropped thumbnails for uniform catalog display.
Build accessibility features
Generate image descriptions for visually impaired users. Provide automatic alt text for images in content management systems.
Try These Prompts
Use the Azure AI Vision Java SDK to generate a caption for the image at local file path /path/to/image.jpg. Show how to set up the client with environment variables and print the caption text with confidence score.
Show me how to use the Azure AI Vision SDK to perform OCR on a scanned document image. Include code to iterate through detected text blocks, lines, and words with their bounding polygons.
Write Java code that analyzes an image using multiple visual features: caption, tags, objects, and read (OCR). Show how to configure the analysis options for gender-neutral captions and specify language.
Demonstrate how to use the async client in Azure AI Vision Java SDK to analyze images. Show the subscription pattern for handling async results and errors.
Best Practices
- Always store Azure credentials in environment variables, never hardcode API keys in source code
- Use async clients for high-volume image processing to improve throughput
- Handle HttpResponseException for robust error handling and retry logic
Avoid
- Do not hardcode API keys or endpoint URLs in your Java source files
- Do not ignore error handling - always catch and log Azure service exceptions
- Do not process images larger than 20MB - resize before analysis to avoid failures