Alpha Matting

Detailed Description

Information Flow algorithm implementaton for alphamatting

Alpha matting is used to extract a foreground object with soft boundaries from a background image.

This module is dedicated to computing alpha matte of objects in images from a given input image and a greyscale trimap image that contains information about the foreground, background and unknown pixels. The unknown pixels are assumed to be a combination of foreground and background pixels. The algorithm uses a combination of multiple carefully defined pixels affinities to estimate the opacity of the foreground pixels in the unkown region.

The implementation is based on [7].

This module was developed by Muskaan Kularia and Sunita Nayak as a project for Google Summer of Code 2019 (GSoC 19).

Functions
void	cv::alphamat::infoFlow (InputArray image, InputArray tmap, OutputArray result)
	Compute alpha matte of an object in an image.

Function Documentation

infoFlow()

void cv::alphamat::infoFlow	(	InputArray	image,
		InputArray	tmap,
		OutputArray	result
	)

#include <opencv2/alphamat.hpp>

Compute alpha matte of an object in an image.

Parameters

image	Input RGB image
tmap	Input greyscale trimap image
result	Output alpha matte image

The function infoFlow performs alpha matting on a RGB image using a greyscale trimap image, and outputs a greyscale alpha matte image. The output alpha matte can be used to softly extract the foreground object from a background image. Examples can be found in the samples directory.