oximedia-cv

Computer vision module for the OxiMedia multimedia framework.

Part of the oximedia workspace — a comprehensive pure-Rust media processing framework.

Overview

oximedia-cv provides computer vision algorithms and image processing capabilities:

• Image Processing: Resize, color conversion, filtering, edge detection
• Detection: Face detection, motion detection, corner detection, YOLO object detection
• Transforms: Affine and perspective transformations
• Enhancement: Super-resolution and denoising (with ONNX)
• Tracking: Optical flow, feature tracking, object tracking (KCF, CSRT, MOSSE, MedianFlow)
• Stabilization: Video stabilization with motion smoothing
• Scene Detection: Video scene changes and shot boundary detection
• Quality Metrics: PSNR, SSIM, VMAF, temporal quality assessment
• Frame Interpolation: Optical flow-based frame rate conversion
• Chroma Key: Green/blue screen with spill suppression and compositing
• Content-aware Scaling: Seam carving and saliency-guided resizing
• Interlace Detection: Interlacing and telecine detection with IVTC recommendations
• Motion Blur: Synthesis and removal via deconvolution
• Fingerprinting: Perceptual content fingerprinting for identification
• Depth Estimation: Monocular depth estimation
• Pose Estimation: Human pose keypoint detection
• Segmentation: Image/video segmentation
• Superpixel: SLIC and compact superpixel algorithms
• Morphology: Erosion, dilation, open/close operations
• Lane Detection: Road lane detection

Features

Image Processing

Detection

Tracking

Enhancement

Usage

use oximedia_cv::image::{ResizeMethod, ColorSpace};
use oximedia_cv::detect::BoundingBox;

// Create a bounding box
let bbox = BoundingBox::new(10.0, 20.0, 100.0, 150.0);
assert!(bbox.area() > 0.0);

Module Structure

src/
├── lib.rs                  # Crate root with re-exports
├── error.rs                # CvError and CvResult
├── bounding_box.rs         # Bounding box types
├── image/                  # Image processing operations
├── detect/                 # Detection algorithms (face, motion, YOLO)
├── transform/              # Geometric transformations
├── enhance/                # Enhancement (super-resolution, denoising, CPU upscale)
├── tracking/               # Object tracking (KCF, CSRT, MOSSE, MedianFlow)
├── stabilize/              # Video stabilization
├── scene/                  # Scene and shot boundary detection
├── quality/                # Quality metrics (PSNR, SSIM, temporal)
├── interpolate/            # Frame interpolation
├── chroma_key/             # Chroma keying and compositing
├── scale/                  # Content-aware scaling
├── interlace/              # Interlace/telecine detection
├── motion_blur/            # Motion blur synthesis and removal
├── fingerprint/            # Perceptual fingerprinting
├── depth_estimation.rs     # Depth estimation
├── pose_estimation.rs      # Pose estimation
├── segmentation.rs         # Image segmentation
├── superpixel.rs           # Superpixel algorithms
├── morphology.rs           # Morphological operations
├── contour.rs              # Contour detection
├── feature_extract.rs      # Feature extraction
├── feature_match.rs        # Feature matching
├── keypoint.rs             # Keypoint management
├── motion_vector.rs        # Motion vector computation
├── optical_flow_field.rs   # Optical flow field
├── hough_transform.rs      # Hough transform
├── histogram_backproject.rs # Histogram back-projection
├── color_cluster.rs        # Color clustering
├── obj_tracking.rs         # Object tracking orchestration
├── texture_analysis.rs     # Texture analysis
└── ml/                     # Machine learning (ONNX, feature-gated)

Feature Flags

[dependencies]
oximedia-cv = { version = "0.1.3", features = ["onnx"] }

License

Apache-2.0 — Copyright 2024-2026 COOLJAPAN OU (Team Kitasan)