Image processing of objects and a background

Patent metadata

Field	Value
Publication number	US-10019781-B2
Application number	US-201615267132-A
Country	US
Kind code	B2
Filing date	Sep 15, 2016
Priority date	Sep 15, 2016
Publication date	Jul 10, 2018
Grant date	Jul 10, 2018

Abstract

Disclosed aspects relate to image processing. An image includes one or more original objects and an original background. The image has a first resolution. A first object having a second resolution is obtained. The second resolution is not higher than the first resolution. The first object is corresponds to at least one of the one or more original objects. A first background with a third resolution is obtained. The third resolution is lower than the second resolution. The first background corresponds to a portion of the image which includes at least the original background. The first object and the first background are sent to a user to display a processed image.

First claim

What is claimed is:

1. A computer-implemented method for image processing, wherein an image includes one or more original objects and an original background, and wherein the image has a first resolution, the method comprising:
identifying one or more original objects from the image;
performing Natural Language Processing (NLP) to a context associated with the image to get an output indicating a subject of the context associated with the image;
selecting, according to the subject, a first object from the one or more identified original objects, wherein the selecting includes matching a label that corresponds to one of the one or more identified original objects with the subject according to the output and determining that the one of the one or more identified original objects corresponds to the matched label as the first object;
estimating a time to load the image by a user;
performing, in response to the estimated time being longer than a threshold, a compression of the first object, wherein the compressed first object has a second resolution not higher than the first resolution;
obtaining a first background having a third resolution, wherein the first background corresponds to a portion of the image which includes at least the original background, and wherein the third resolution is lower than the second resolution; and
sending the compressed first object and the first background to the user to display a processed image.

2. The method of claim 1 , wherein the first background is a low resolution image obtained by compressing a selection from the group consisting of: the image, and a remaining part of the image which excludes the original object corresponding to the first object.

3. The method of claim 1 , further comprising:
in response to a second object different from the first object being requested by the user, obtaining the second object, wherein the second object corresponds to at least one of the one or more original objects, wherein the second object has a fourth resolution, and wherein the fourth resolution is not higher than the first resolution and is higher than the third resolution; and
sending the second object to the user.

4. The method of claim 3 , further comprising:
providing a second label corresponding to the second object to the user; and
in response to a second object different from the first object being requested by the user, obtaining the second object includes: in response to a user selection of a selected label, obtaining a second object which corresponds to the selected label.

5. The method of claim 1 , wherein the third resolution is lower than the second resolution by a threshold amount.

6. A computer program product being tangibly stored on a non-transient machine-readable medium and comprising machine-executable instructions, the instructions, when executed on a device for image processing, wherein an image includes one or more original objects and an original background, and wherein the image has a first resolution, causing the device to:
perform Natural Language Processing (NLP) to a context associated with the image to get an output indicating a subject of the context;
identify one or more original objects from the image;
select, based on the identified one or more original objects and based on the NLP, a first object from the one or more identified original objects corresponding to the subject of the context, wherein the selecting includes matching a label that corresponds to one of the one or more identified original objects with the subject according to the output and determining that the one of the one or more identified original objects corresponds to the matched label as the first object;
estimate a time to load the image by a user;
perform, based on the estimated time being longer than a threshold, a compression of the first object, wherein the compressed first object has a second resolution that is not higher than the first resolution;
obtain a first background having a third resolution, wherein the first background corresponds to a portion of the image having at least the original background, and wherein the third resolution is lower than the second resolution; and
send the compressed first object and the first background to the user to display a processed image.

7. The computer program product of claim 6 , wherein
the first background is a low resolution image obtained by compressing a selection from the group consisting of: the image, and a remaining part of the image which excludes the original object corresponding to the first object.

8. The computer program product of claim 6 , wherein the instructions, when executed on the device, further cause the device to:
in response to a second object different from the first object being requested by the user, obtain the second object, wherein the second object corresponds to at least one of the one or more original objects, the second object has a fourth resolution and the fourth resolution is not higher than the first resolution and is higher than the third resolution; and
send the second object to the user.

9. A computing system, the system comprising:
a memory storing instructions thereon for image processing, wherein an image includes one or more original objects and an original background, wherein the image has a first resolution; and
a processor, the processor communicatively coupled to the memory, based on the stored instructions the processor configured to:
identify one or more original objects from the image;
perform Natural Language Processing (NLP) to a context associated with the image to get an output indicating a subject of the context associated with the image;
select, according to the subject, a first object from the one or more identified original objects, wherein the selecting includes matching a label that corresponds to one of the one or more identified original objects with the subject according to the output and determining that the one of the one or more identified original objects corresponds to the matched label as the first object,
wherein the first object has a second resolution, and wherein the second resolution is not higher than the first resolution;
estimate a time to load the image by a user according to a user network condition and a size of the image;
perform, in response to the estimated time being longer than a threshold, a compression of a first background;
obtain the first background having a third resolution, wherein the first background corresponds to a portion of the image including at least the original background, and wherein the third resolution is lower than the second resolution; and
send the first object and the first compressed background to the user to display a processed image.

10. The computing system of claim 9 , further comprising:
in response to a second object different from the first object being requested by the user, obtaining the second object, wherein the second object corresponds to at least one of the one or more original objects, wherein the second object has a fourth resolution, and wherein the fourth resolution is not higher than the first resolution and is higher than the third resolution; and
sending the second object to the user.

Assignees

• IBM

Inventors

• Chen Hsiao-Yung
• Huang chao yuan
• Lai Hung-Jung
• Qian Yin
• Shen Fei

Classifications

• G06K9/726

  Physics · mapped topic

• G06T2207/20144

  Physics · mapped topic

• G06T2207/20148

  Physics · mapped topic

• G06T7/0081

  Physics · mapped topic

• G06T3/4092 Primary

  Image resolution transcoding, e.g. by using client-server architectures · CPC title