Optimization-based intrinsic image extraction method
An extraction method and eigenmap technology, which are applied in the field of intrinsic image extraction based on optimization and intrinsic image extraction, and can solve the problems of inability to process image areas containing black and white textures, unfavorable promotion and application, and poor image adaptability.
Inactive Publication Date: 2012-10-24
BEIJING INSTITUTE OF TECHNOLOGYGY
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] Extraction method of multiple images: extract the intrinsic map image through the image sequence of the same scene under different illumination. This method needs to input multiple images of the same scene, so it is not conducive to popularization and application
[0005] Learning-based extraction methods: Extracting intrinsic maps by learning a classifier from an image database, this method requires a large number of images as learning samples, but the resulting model is not well adaptable to a wide variety of images in nature
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0047] The implementation method of the present invention will be described in detail below in conjunction with the accompanying drawings.
[0048] figure 1 A flow chart of the optimization-based eigenmap extraction method of this embodiment is given, and its main steps are as follows:
[0049] 1. Calculation of local domain weights:
[0050] The size of the field is set to 3*3, for each pixel of the input image I, a field with a size of 3*3 centered on the pixel is selected, according to and Calculate the weight of each other pixel in the domain, where express and angle, and is unitized I i and I j The resulting vector, Indicates the variance of the angle between other pixels in the field of pixel i and i, Indicates the variance of the Y values of all pixels in the area of pixel i. The parameter β is used to control the influence of the included angle on the weight. The user can set it to an integer, but it is found in the experiment that when β is a sma...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention relates to an optimization-based intrinsic image extraction method, comprising the following steps of: 1, setting a domain size N; defining a weight function, representing the albedo of each pixel point in the image as the albedo weighted sum of other pixel points in the domain taking the pixel point as the center; solving equations set formed by the albedo of each pixel point, and acquiring the values of the albedo of all the pixel points in the image; 2, defining an energy equation, and building an optimization problem; and 3, using a Gauss-Seide iterative method to solve the optimization problem so as to acquire an albedo intrinsic image R and an illumination intrinsic image s. compared with the prior art, the intrinsic image obtained by the method has higher technical quality, and solves the problem that the interactive intrinsic image extraction method in the prior art cannot process the black and white vein domains.
Description
technical field [0001] The invention relates to an intrinsic image extraction method, in particular to an optimization-based intrinsic image extraction method, which belongs to the field of computer vision and pattern recognition. Background technique [0002] The surface color of an object observed by the human eye is affected by many factors, including the shape and material of the object, the position and color of the light source, and the position of the observer. Albedo and illumination are the two most important influencing factors. Albedo describes The reflection characteristics of an object itself to incident light reflect the true color of the object, and the illumination describes the light information projected on the surface of the object, reflecting the light and dark changes on the surface of the object. The intrinsic image extraction method is to extract the albedo information and illumination information in the image, which are called albedo intrinsic map and...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/46
Inventor 沈建冰杨小汕
Owner BEIJING INSTITUTE OF TECHNOLOGYGY