Using the SDK

The Perfectly Clear SDK provides two API options for integrating image correction into your applications. Choose the API that best fits your development environment and requirements.

Available APIs

The SDK includes a native C API, a .NET wrapper for C# applications, and a JNI wrapper for Java. All APIs provide access to the same underlying correction engine and produce identical results.

Processing workflow

Both APIs follow the same general workflow for image correction:

1. Initialize — Set up license protection and create an engine instance
2. Load AI models — Load scene detection and AI correction models (optional but recommended)
3. Load image — Read image data into memory
4. Analyze — Calculate correction profile for the image
5. Apply — Apply corrections using the profile and parameters
6. Save — Write the corrected image to disk or memory

The SDK supports two approaches: a simplified single-function call (AutoCorrect) or a two-step process (Calc + Apply) that allows inspection and modification of parameters between analysis and correction.

Choosing an approach

For most batch processing scenarios, use AutoCorrect with AI scene detection enabled. This provides optimal corrections with minimal code.

For interactive applications where users adjust parameters, use the two-step Calc + Apply approach. This lets you analyze the image once, then apply corrections multiple times with different settings without re-analyzing.

Next steps

Explore more on the architecture, core objects and API patterns: