Computer implemented system and method for image texture determination

Patent metadata

Field	Value
Publication number	US-2016284100-A1
Application number	US-201615019794-A
Country	US
Kind code	A1
Filing date	Feb 9, 2016
Priority date	Mar 25, 2015
Publication date	Sep 29, 2016
Grant date	—

Abstract

This disclosure relates generally to image processing, and more particularly to image texture determination. In one embodiment a processor a memory coupled to the processor, wherein the processor coupled with a plurality of modules stored in the memory: At least one image having a plurality of pixels is accepted. Any noise is removed from the image to obtain at least one noise free image. The at least one noise free image is converted to at least one gray scale image. Horizontal and Vertical Gradient for plurality of pixels of the at least one gray scale image are computed. Gradient magnitude is calculated for the generated gradient. Histogram of the gradient magnitude is generated based on the gradient magnitude, and the plurality of generated histograms are compared with a plurality of predetermined histograms.

First claim

What is claimed is:

1 . A computer implemented system ( 100 ) for determining image texture, said system ( 100 ) comprising:
a processor;
a memory coupled to the processor, wherein the processor coupled with a plurality of modules stored in the memory, and wherein the plurality of modules 208 comprising:
an input module ( 30 ) accepts at least one image having a plurality of pixels;
a noise removal module ( 40 ) receives said at least one image and further configured to remove noise from the at least one image to obtain at least one noise free image;
a gray scale converter module ( 50 ) receives the at least one noise free image and further configured to convert the at least one noise free image into at least one gray scale image;
a gradient module ( 60 ) receives at least one gray scale image and further configured to compute a gradient in horizontal and vertical direction for a plurality of pixels of the at least one gray scale image;
a magnitude determiner ( 70 ) receives said gradient in horizontal and vertical direction for said plurality of pixels of the at least one gray scale image and further configured to compute a gradient magnitude for the plurality of pixels of the at least one gray scale image;
a histogram generator module ( 80 ) receives said gradient magnitude for the plurality of pixels of said at least one gray scale image and further configured to generate a histogram of gradient magnitude for said at least one gray scale image based on said received gradient magnitude; and
a texture determiner module ( 90 ) having a comparator ( 92 ) receives said plurality of predetermined histogram and the histogram generator module ( 80 ) receives said generated histogram of gradient magnitude for the said gray scale image and further configured to compare the generated histogram of gradient magnitude with a plurality of predetermined histograms to determine the image texture.

2 . The system as claimed in claim 1 , wherein the noise removal module ( 40 ) is configured to use Gaussian blur technique to remove noise from the at least one image.

3 . The system as claimed in claim 1 , wherein the generated histogram of gradient magnitude is independent of rotation of the at least one image.

4 . A computer implemented method for determining image texture, said method comprising:
accepting, by a processor, a at least one image having a plurality of pixels; removing, by the processor, noise from the at least one image to obtain at least one noise free image; converting, by the processor, the noise free image into at least one gray scale image; computing a gradient in horizontal and vertical direction for a plurality of pixels of the at least one gray scale image; computing by the processor, a gradient magnitude for the at least one gray scale image; generating by the processor, a histogram of gradient magnitude for the at least one gray scale image; and comparing by the processor, the histogram of gradient magnitude with a plurality of predetermined histograms to determine the image texture.

5 . The method as claimed in claim 4 , wherein the computed histogram of gradient magnitude is rotational invariant.

6 . The method as claimed in claim 4 , further includes storing the plurality of predetermined histograms.

7 . A non-transitory computer-readable medium having embodied thereon a computer program for executing a method for determining image texture, the method comprising:
accepting a plurality of images having a plurality of pixels; removing noise from the at least one image to obtain at least one noise free image; converting the noise free image into at least one gray scale image; computing a gradient in horizontal and vertical direction for a plurality of pixels of the at least one gray scale image; computing a gradient magnitude for the at least one gray scale image; generating a histogram of gradient magnitude for the at least one gray scale image; and comparing the generated histogram of gradient magnitude with a plurality of predetermined histograms to determine the image texture.

8 . A computer implemented method for determining image texture, said method comprising:
accepting an image having a plurality of pixels; partitioning the image into a plurality of image segments; generating a plurality of histograms of gradient magnitudes for the plurality of image segments, wherein generating an HGM for an image segment of the plurality of image segment comprises:
removing noise from the image segment to obtain at least one noise free image segment,
converting said noise free image segment into a gray scale image segment,
computing a gradient in horizontal and vertical direction for a plurality of pixels of the gray scale image segment,
computing a gradient magnitude for the gray scale image segment, and
generating the histogram of gradient magnitude for the gray scale image segment based on the gradient magnitude for the gray scale image segment;
comparing the plurality of histograms of gradient magnitude with the plurality of predetermined histograms to determine the texture of the image segment; and clustering the plurality of HGMs into a plurality of clusters based on the comparison, wherein each cluster of the plurality of clusters is associated with a distinct texture.

Assignees

• Tata Consultancy Services Ltd

Inventors

• Sharma Monika
• Ghosh Hiranmay

Classifications

• G06T7/40 Primary

  Analysis of texture (depth or shape recovery from texture G06T7/529) · CPC title

• G06V10/54 Primary

  relating to texture · CPC title

• G06V10/50

  by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis · CPC title

• G06V10/758

  Involving statistics of pixels or of feature values, e.g. histogram matching · CPC title

• G06T5/002

  Physics · mapped topic