Detector

A collection of detector implementations for images and signal streams. More...

Collaboration diagram for Detector:

Modules
	Detect Activity
	Interfaces for detecting activity semantics from time-varying signals.
	Detect Background
	Interfaces for detecting when signal matches a background model.
	Detect Foreground
	Interfaces for detecting when signal matches a foreground model.

Namespaces
	fromState
	Detector based on state vector input.
	inImage
	Detector based on image input.
	inImageRGBD
	Detector based on RGBD imagery.

Classes
class	ActivityState
	Basic version of an activity state. More...
class	Base
	Base or root implementation of detector class. More...
class	DetectorState
	Basic version of a detector state. More...
class	fgAppearance
	Appearance based object detection. More...
class	fgDifferences
	Image differences foreground model. More...
class	fgGaussian
	Single Gaussian target/foreground model with diagonal covariance. More...
class	fgImage
	The most basic object or instance detector from image input. More...
class	fromState
	Basic detector from state vector input. Should be overloaded, but simple detectors can be created with the right pre/post processor. More...
class	inImage
	The most basic object or instance detector from image input. More...
class	inImageRGBD
	Most basic object or instance detector from RGBD image input. More...
class	targetCorner
	Class instance for extracting targets that lie with the corner region of the RGB color cube, or the extremal regions of a color-space based on a planar cut. More...
class	targetMagenta
	Class instance for extracting bright/neon magenta targets, which tend to lie in a corner region of the RGB color cube. More...
class	targetNeon
	Detection interface for neon colored targets. More...
class	targetSG
	Single-Gaussian based target detection with full covariance. More...

Functions
def	__init__ (self, config=CfgMoving(), processor=None)
	Simple activity detector based on motion/velocity. More...

Detailed Description

A collection of detector implementations for images and signal streams.

Attempts to encapsulate the generic functionality of a detector through this package. Here, detector and recognition are loosely identified. What is detected can vary: objects, targets, actions, activities, etc. Basically anything that can be given semantic meaning from a signal. Not all semantic labels need be from detectors though. Some are derivec elsewhere or downstream of the base detector process.

Function Documentation

__init__()

def __init__	(		self,
			processor = CfgMoving(),
			config = None
	)

Simple activity detector based on motion/velocity.

A motion detector for track signals from an image.

Constructor for isMoving activity instance.

Parameters

[in]

processor

A state processor.

Augments super class with a GONE track state since image-based measurements can be missing when the target of interest leaves the field of view.

Constructor for isMovingInImage activity instance.

Parameters

[in]

processor

A state processor.