Image processing device and image processing program

Patent metadata

Field	Value
Publication number	US-2022180546-A1
Application number	US-202017599441-A
Country	US
Kind code	A1
Filing date	Mar 30, 2020
Priority date	Mar 28, 2019
Publication date	Jun 9, 2022
Grant date	—

Abstract

An image processing device can use a calculation formula based on an ellipse to approximate a base function of a reference GMM. The burden rate according to a co-occurrence correspondence point can be approximately determined by a calculation in which the Manhattan distance to the ellipse and the co-occurrence correspondence point and the width of the ellipse are input to a calculation formula for the burden rate based on the base function. The width of the ellipse is quantized by the nth power of 2 (where n is an integer of 0 or greater), and the calculation can be carried out by means of a bit shift.

First claim

1 . An image processing device comprising:
image acquiring means for acquiring an image; co-occurrence distribution acquiring means for acquiring a distribution of co-occurrences of luminance gradient directions from the acquired image; calculating means for calculating a base function using the distribution of the co-occurrences, and calculating a feature amount of the image using the base function; and outputting means for outputting the calculated feature amount.

2 . The image processing device according to claim 1 , comprising parameter storing means for storing a parameter for defining a base function which approximates a Gaussian mixture model serving as a reference for image recognition,
wherein the calculating means calculates the feature amount of the image using the Gaussian mixture model by substituting a distance from each co-occurrence point constituting the acquired distribution of the co-occurrences to a center of the base function and the stored parameter in the base function formula.

3 . The image processing device according to claim 1 ,
wherein the parameter storing means stores the parameter in accordance with each Gaussian distribution constituting the Gaussian mixture model, and the calculating means calculates a value which becomes an element of the feature amount by using the parameter of the Gaussian distribution in accordance with each Gaussian distribution.

4 . The image processing device according to claim 1 ,
wherein the calculating means approximately calculates a burden rate using the distribution of co-occurrences for each Gaussian distribution as a value of the element of the feature amount.

5 . The image processing device according to claim 1 ,
wherein the parameter is a constant which defines an ellipse corresponding to a width of each of the Gaussian distributions.

6 . The image processing device according to claim 5 ,
wherein a direction of a maximum width of the ellipse is parallel or perpendicular to an orthogonal coordinate axis defining the Gaussian mixture model.

7 . The image processing device according to claim 1 ,
wherein the parameter is quantized to a power of 2, and the calculating means performs the calculation by using a bit shift.

8 . The image processing device according to claim 1 , comprising image recognizing means for performing image recognition of the image by using the feature amount output by the outputting means.

9 . A non-transitory computer-readable storage medium storing a computer-executable program for causing a computer to perform functions comprising:
acquiring an image; acquiring a distribution of co-occurrences of luminance gradient directions from the acquired image; calculating a base function using the distribution of the co-occurrences, and calculating a feature amount of the image using the base function; and outputting the calculated feature amount.

Assignees

• Aisin Corp
• Kyushu Inst Tech

Inventors

• Yamada Hideo
• Shibata Masatoshi
• TAMUKOH Hakaru
• Enokida Shuichi
• Yoshihiro Kazuki

Classifications

• G06T1/60

  Memory management · CPC title

• G06V10/50 Primary

  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

• G06V10/60

  relating to illumination properties, e.g. using a reflectance or lighting model · CPC title

• G06T7/60 Primary

  Analysis of geometric attributes · CPC title