Into comes with an extensive set of built-in functionality for pattern recognition and image analysis applications. The following is an incomplete list of built-in features.

Core functionality

• Plug-in architecture allows features to be loaded and unloaded at run time.
• Standard interfaces make it possible to use any software module with any other software module with ease.
• Automatic parallelization and asynchronous processing utilize all processor cores for best performance.
• Network transparency makes it possible to distribute load across network and to interact directly with web browsers.
• Scriptability makes it possible to build applications with JavaScript.
• Serialization library makes it easy to save and load anything.
• Powerful matrix and linear algebra library forms a reliable basis for more complex algorithms.

Networking

• HTTP client
• HTTP server
• Remote object framework
  • Extremely easy-to-use remote functions
  • Signals and slots over network
  • Effortless integration to web applications

Image and Video Processing

• Image formats
  • Wide variety of image and video file types supported
  • Matrix and line-scan cameras, including network cameras
  • Support for writing images and video files
• Image filtering
  • Linear filtering
  • Median filtering
  • Automatic optimization of separable filters
• Edge detection (Canny, Sobel, Roberts, LoG ...)
• Thresholding techniques
  • One-level and two-level thresholds
  • Automatic threshold calculation
  • Hysteresis thresholding
  • User-defineable thresholding functions
  • Adaptive thresholding
• Binary morphology
  • Opening, closing etc.
  • Hit-and-miss transform
  • Skeletonization
• Labeling
• Histogram processing
  • Learning quantizers
  • Equalization
  • Backprojection
  • Percentiles
  • Multi-dimensional histograms
• Geometric transformations
  • Scaling
  • Rotation
  • Shear
  • User-defined
• Transforms
  • Hough transform
  • 2D FFT
• Color processing
  • Blob matching (ellipsoid-based)
  • Color space conversions (RGB, HSV, XYZ, CIELAB, ...)
  • Color normalization
• Arbitrarily shaped regions-of-interest
• Matching
  • Feature point based matching framework
  • Correlation based template matching
  • Shape matching
• Tracking
  • Background modeling
  • User-extensible tracking framework
• Camera Calibration
• Stereo imaging with two or more cameras
• OCR, handwritten digit recognition
• OpenCV integration

Pattern Recognition

• Vector quantization framework
• Kernel methods framework
• Nearest neighbor classifier
• k-NN classifier
• k-d trees
• Perceptron
• Kernel perceptron
• Kernel adatron
• SOM
• Boosted classifiers
• Rule-based classifier
• Different distance measures and kernel functions
• Confusion matrix
• Texture feature extraction
  • LBP
  • Wavelet energy
  • FFT-based features
• Color and gray-level feature extraction
• Shape features

Other

• DSP
  • Convolution and correlation
  • FFT (1D and 2D)
  • Wavelet transforms (1D and 2D)
• Optimization framework
  • BFGS
  • Levenberg-Marquardt
  • RANSAC
  • Linear assignment problem
• SQL database interfaces
• GUI building blocks
  • Image display
  • Visual training of SOM classifiers
  • Dynamically loading image list
  • Gauges