Crate vil_vision 

VIL Image Understanding (I04)

Infrastructure and traits for image analysis, object detection, OCR, and image embedding. This crate provides trait definitions and format detection. Actual vision backends plug in via the ImageAnalyzer and ImageEmbedder traits.

Quick Start

use vil_vision::{ImageFormat, detect_format, VisionConfig};

let png_header = vec![0x89, 0x50, 0x4E, 0x47, 0x0D, 0x0A, 0x1A, 0x0A];
assert_eq!(detect_format(&png_header), ImageFormat::Png);

let config = VisionConfig::new().ocr_enabled(true).min_confidence(0.7);

Re-exports

pub use analyzer::DetectedObject;

pub use analyzer::ImageAnalysis;

pub use analyzer::ImageAnalyzer;

pub use analyzer::NoopAnalyzer;

pub use analyzer::VisionError;

pub use config::VisionConfig;

pub use embedding::ImageEmbedder;

pub use embedding::NoopEmbedder;

pub use format::detect_format;

pub use format::ImageFormat;

pub use plugin::VisionPlugin;

pub use semantic::VisionEvent;

pub use semantic::VisionFault;

pub use semantic::VisionFaultType;

pub use semantic::VisionState;

Modules

analyzer

config

embedding

format

handlers

VIL pattern HTTP handlers for the vision plugin.

pipeline_sse

SSE pipeline builders for vision analysis operations.

plugin

VilPlugin implementation for image analysis integration.

semantic

Semantic types for image analysis operations.