Image processing method, device and medium

Patent metadata

Field	Value
Publication number	US-10534972-B2
Application number	US-201615368663-A
Country	US
Kind code	B2
Filing date	Dec 4, 2016
Priority date	Dec 4, 2015
Publication date	Jan 14, 2020
Grant date	Jan 14, 2020

Abstract

An image processing method, device and medium are provided. The method includes: acquiring candidate areas from an image to be processed, each of the candidate areas including a reference target; extracting a predetermined characteristic of each of the candidate areas; calculating an evaluation value of each of the candidate areas according to the predetermined characteristic; and acquiring a snapshot of the image to be processed according to the evaluation values.

First claim

What is claimed is:

1. An image processing method, comprising:
acquiring candidate areas from an image to be processed, each of the candidate areas comprising a reference target;
extracting a predetermined characteristic of each of the candidate areas, wherein the predetermined characteristic comprises at least one of a color characteristic, a grain characteristic, a shape characteristic and a space relationship characteristic;
calculating an evaluation value of each of the candidate areas according to the predetermined characteristic; and
acquiring a snapshot of the image to be processed according to the evaluation values,
wherein acquiring candidate areas from an image to be processed comprises:
acquiring the reference target of the image to be processed; and
extracting the candidate area comprising the reference target from the image to be processed according to a predetermined condition, wherein the predetermined condition comprises: if the candidate area comprises a portion of a remaining target other than the reference target, the candidate area is adjusted to comprise entirety of the remaining target; and

wherein calculating an evaluatiojn value of each of the candidate areas according to the predetermined characteristic comprises:
calculating the evaluation value of the candidate area according to a predetermined image evaluating model and the predetermined characteristic extracted from the candidate area, wherein the predetermined image evaluating model is a model between predetermined characteristics and evaluation values and is acquired by training according to predetermined characteristics extracted from sample images and evaluation values of the sample images.

2. The method according to claim 1 , wherein acquiring the reference target of the image to be processed comprises:
performing a target detection on the image to be processed to acquire a target set;
acquiring a click position on the image to be processed from a user; and
determining a target corresponding to the click position in the target set as the reference target.

3. The method according to claim 1 , wherein acquiring the reference target of the image to be processed comprises:
performing a saliency detection on the image to be processed; and
determining a target with a highest saliency as the reference target.

4. The method according to claim 1 , wherein the predetermined condition further comprises:
the reference target is located at a preset position in the candidate area.

5. The method according to claim 1 , wherein acquiring candidate areas from an image to be processed comprises:
acquiring a selected area in the image to be processed from the user, the selected area comprising the reference target; and
determining the candidate areas according to the selected area.

6. The method according to claim 1 , wherein extracting a predetermined characteristic of each of the candidate areas comprises:
extracting a predetermined characteristic of the reference target in the candidate area; and
extracting the predetermined characteristic of the candidate area.

7. The method according to claim 1 , wherein acquiring a snapshot of the image to be processed according to the evaluation values comprises:
acquiring an candidate area with a highest evaluation value; and
determining the candidate area with the highest evaluation value as the snapshot of the image to be processed.

8. The method according to claim 1 , wherein acquiring a snapshot of the image to be processed according to the evaluation values comprises:
ranking the candidate areas according to the evaluation values;
displaying the candidate areas ranked; and
determining a selected candidate area as the snapshot of the image to be processed, according to a selection operation on the candidate areas from the user.

9. The method according to claim 1 , further comprising:
acquiring the sample images;
acquiring a saliency area of each of the sample images;
extracting a first predetermined characteristic of the saliency area and a second predetermined characteristic of a whole sample image;
acquiring a preset evaluation value of the sample image; and
acquiring the predetermined image evaluating model by performing a model training according to the preset evaluation values of the sample image and the predetermined characteristics extracted.

10. The method according to claim 1 , wherein the predetermined characteristic comprises:
a photograph characteristic of an image comprising at least one of an aperture when photographing, a shutter, a white balance, a photosensibility, a focus, photograph time, a photograph condition, a camera brand, a version, a color code, a voice recorded when photographing, a photograph position, and a thumbnail.

11. An image processing device, comprising:
a processor; and
a memory configured to store instructions executable by the processor;
wherein the processor is configured to:
acquire candidate areas from an image to be processed, each of the candidate areas comprising a reference target;
extract a predetermined characteristic of each of the candidate areas, wherein the predetermined characteristic comprises at least one of a color characteristic, a grain characteristic, a shape characteristic and a space relationship characteristic;
calculate an evaluation value of each of the candidate areas according to the predetermined characteristic; and
acquire a snapshot of the image to be processed according to the evaluation values,
wherein the processor configured to acquire candidate areas from an image to be processed is further configured to:
acquire the reference target of the image to be processed; and
extract the candidate area comprising the reference target from the image to be processed according to a predetermined condition, wherein the predetermined condition comprises: if the candidate area comprises a portion of a remaining target other than the reference target, the candidate area is adjusted to comprise entirety of the remaining target; and

wherein calculating an evaluation value of each of the candidate areas according to the predetermined characteristic comprises:
calculating the evaluation value of the candidate area according to a predetermined image evaluating model and the predetermined characteristic extracted from the candidate area, wherein the predetermined image evaluating model is a model between predetermined characteristics and evaluation values and is acquired by training according to predetermined characteristics extracted from sample images and evaluation values of the sample images.

12. The image processing device according to claim 11 , wherein the processor configured to acquire the reference target of the image to be processed is further configured to:
perform a target detection on the image to be processed to acquire a target set;
acquire a click position on the image to be processed from a user; and
determine a target corresponding to the click position in the target set as the reference target.

13. The image processing device according to claim 11 , wherein the processor configured to acquire the reference target of the image to be processed is further configured to:
perform a saliency detection on the image to be processed; and
determine a target with a highest saliency as the reference target.

14. The image processing device according to claim 11 , wherein the predetermined condition further comprises:
the reference target is located at a preset position in the candidate area.

15. The image processing device according to claim 11 , wherein th

Assignees

• Xiaomi Inc

Inventors

• Wang Baichao
• Qin Qiuping
• Hou Wendi

Classifications

• G06V10/235 Primary

  based on user input or interaction · CPC title

• G06F18/2193 Primary

  based on specific statistical tests · CPC title

• H04N23/60

  Control of cameras or camera modules · CPC title

• H04N23/64

  Computer-aided capture of images, e.g. transfer from script file into camera, check of taken image quality, advice or proposal for image composition or decision on when to take image · CPC title

• G06F3/0488

  using a touch-screen or digitiser, e.g. input of commands through traced gestures · CPC title