Method and apparatus of determining air quality

Patent metadata

Field	Value
Publication number	US-9740967-B2
Application number	US-201615018005-A
Country	US
Kind code	B2
Filing date	Feb 8, 2016
Priority date	Oct 31, 2013
Publication date	Aug 22, 2017
Grant date	Aug 22, 2017

Abstract

In an approach to determining an air quality based on image, a computer performs a method, the method comprising: determining an image to be measured; determining an air quality model of at least one key area, the at least one key area including a key area associated with the image to be measured; acquiring a reference clear image corresponding to the image to be measured; and determining an air quality index to be measured based on an air quality related feature extracted from the image to be measured, the corresponding reference clear image, and the determined air quality model of the key area associated with the image to be measured.

First claim

The invention claimed is:

1. A method of determining an air quality in real-time at a location from a mobile terminal image, the method comprising:
determining, by one or more computer processors, an image to be measured, wherein the image is received from a user of a mobile terminal with a camera device;
determining, by one or more computer processors, an air quality model of at least one key area, the at least one key area including a key area associated with the image to be measured;
acquiring, by one or more computer processors, a reference clear image corresponding to the image to be measured; and
determining, by one or more computer processors, an air quality index of a location to be measured based on an air quality related feature extracted from the image to be measured, the corresponding reference clear image, and the determined air quality model of the key area associated with the image to be measured.

2. The method according to claim 1 , wherein the air quality related feature comprises at least one of: a luminance, a texture and a density gradient.

3. The method according to claim 1 , wherein determining an air quality model of at least one key area comprises:
determining, by one or more computer processors, the at least one key area;
acquiring, by one or more computer processors, a reference clear image, a training image under poor air quality and corresponding actual air quality index in at least one location of the at least one key area; and
training, by one or more computer processors, an air quality model of the at least one key area based on an air quality related feature extracted from the reference clear image and the training image and based on the actual air quality index.

4. The method according to claim 1 , wherein, acquiring a reference clear image corresponding to the image to be measured comprises: determining a first reference clear image based on the location of the image to be measured, the first reference clear image is a reference clear image in an associated key area closest to the image to be measured; and
wherein determining the air quality index to be measured based on the air quality feature extracted from the image to be measured, the corresponding reference clear image and the determined air quality model comprises: determining the air quality index of the location to be measured at least based on the first reference clear image.

5. The method according to claim 1 , wherein, acquiring a reference clear image corresponding to the image to be measured comprises: determining a second reference clear image by performing scene matching based on at least one feature of the image to be measured, the second reference clear image is a reference clear image in the associated key area best matched with the image to be measured; and
wherein determining the air quality index of the location to be measured based on the air quality feature extracted from the image to be measured, the corresponding reference clear image and the determined air quality model of the key area comprises: determining the air quality index of the location to be measured at least based on the second reference clear image.

6. The method according to claim 1 , wherein, acquiring a reference clear image corresponding to the image to be measured comprises: determining a third reference clear image by performing haze removal on the image to be measured; and
wherein determining the air quality index of the location to be measured based on the air quality feature extracted from the image to be measured, the corresponding reference clear image and the determined air quality model of the key area comprises: determining the air quality index of the location to be measured at least based on the third reference clear image.

7. The method according to claim 1 , further comprising:
acquiring, by one or more computer processors, additional information of the image to be measured, the additional information comprises at least one of: shooting parameter information of the image to be measured and geographic location information of the image to be measured; and
wherein acquiring a reference clear image corresponding to the image to be measured comprises: acquiring, by one or more computer processors, a reference clear image corresponding to the image to be measured based on the additional information.

8. A computer system of determining an air quality in real-time at a location from a mobile terminal image, the computer system comprising:
one or more computer processors;
one or more computer readable storage devices; and
program instructions stored on the one or more computer readable storage devices for execution by at least one of the one or more computer processors, the stored program instructions comprising:
program instructions to determine an image to be measured, wherein the image is received from a user of a mobile terminal with a camera device;
program instructions to determine an air quality model of at least one key area, the at least one key area including a key area associated with the image to be measured;
program instructions to a reference clear image corresponding to the image to be measured; and
program instructions to determine an air quality index of a location to be measured based on an air quality related feature extracted from the image to be measured, the corresponding reference clear image, and the determined air quality model of the key area associated with the image to be measured.

9. The computer system according to claim 8 , wherein the air quality related feature comprises at least one of: a luminance, a texture and a density gradient.

10. The computer system according to claim 8 , further comprising, for an image to be measured:
program instructions to determine the at least one key area;
program instructions to a reference clear image, a training image under poor air quality and corresponding actual air quality index in at least one location of the at least one key area; and
program instructions to train an air quality model of the at least one key area based on an air quality related feature extracted from the reference clear image and the training image and based on the actual air quality index.

11. The computer system according to claim 8 , wherein,
program instructions to acquire a reference clear image corresponding to the image to be measured comprise: program instructions to determine a first reference clear image based on the location of the image to be measured, the first reference clear image is a reference clear image in an associated key area closest to the image to be measured; and
wherein the program instructions to determine the air quality index of the location to be measured based on the air quality feature extracted from the image to be measured, the corresponding reference clear image and the determined air quality model comprise: program instructions to determine the air quality index of the location to be measured at least based on the first reference clear image.

12. The computer system according to claim 8 , wherein,
program instructions to acquire a reference clear image corresponding to the image to be measured comprise: program instructions to determine a second reference clear image by performing scene matching based on at least one feature of the image to be measured, the second reference clear image is a reference clear image in the associated key area best matched with the image to be measured; and
wherein the program instructions to determine the air quality index of the location to be measured based on the air quality feature extracted from the image to be measured, the corresponding reference clear image a

Assignees

• IBM

Inventors

• Gong Min
• Wang Yu
• Wang Zhi H
• Yan Junchi
• Zhang Chao
• Zhao Qian K
• Zhu Jun

Classifications

• G01N33/0062 Primary

  concerning the measuring method or the display, e.g. intermittent measurement or digital display · CPC title

• G06F18/214

  Generating training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title

• G06K9/66 Primary

  Physics · mapped topic

• G06K9/00

  Physics · mapped topic

• G01J1/42

  using electric radiation detectors (optical or mechanical part G01J1/04; by comparison with a reference light or electric value G01J1/10) · CPC title