Method, apparatus and device for evaluating image tracking effectiveness and readable storage medium

Patent metadata

Field	Value
Publication number	US-11023781-B2
Application number	US-201916510164-A
Country	US
Kind code	B2
Filing date	Jul 12, 2019
Priority date	Aug 3, 2018
Publication date	Jun 1, 2021
Grant date	Jun 1, 2021

Abstract

The present disclosure provides a method, an apparatus, a device for evaluating image tracking effectiveness and a readable storage medium. The method includes: extracting feature points from a target image according to a predefined algorithm; determining first texture information for the target image according to the feature points; processing the target image according to a predefined processing policy to derive a contrast image, and determining second texture information for the target image according to the contrast image; and rating the target image according to the first texture information and the second texture information. In the solution provide by the present disclosure, the tracking effectiveness for the target image can be known without having to place the target image into any actual product, nor having to wait until the product has been finalized before the target image tracking effectiveness is determined. Moreover, efficiency can be enhanced when evaluating image tracking effectiveness.

First claim

What is claimed is:

1. A method for evaluating image tracking effectiveness, comprising:
extracting feature points from a target image according to a predefined algorithm;
determining first texture information for the target image according to the feature points;
processing the target image according to a predefined processing policy to derive a contrast image, and determining second texture information for the target image according to the contrast image; and
rating tracking effectiveness for the target image according to the first texture information and the second texture information;
wherein the determining first texture information for the target image according to the feature points comprises:
determining a first texture rating for the target image according to the feature points, wherein the first texture rating comprises at least one of the following texture ratings:
a texture richness rating, a texture distribution rating, and a texture repetition rating;
the determining second texture information for the target image according to the contrast image comprises:
determining a texture contrast rating for the target image according to the contrast image;
correspondingly, the rating tracking effectiveness for the target image according to the first texture information and the second texture information comprises:
determining a lowest rating among the first texture rating and the texture contrast rating as a rating for the target image.

2. The method according to claim 1 , wherein the determining a texture richness rating for the target image according to the feature points comprises:
deriving a quantity of the feature points; and determining the texture richness rating according to the quantity.

3. The method according to claim 1 , wherein the determining a texture distribution rating for the target image according to the feature points comprises:
equally partitioning the target image into a number of grids;
deriving a quantity of the feature points in each of the grids, and determining the grids in which the quantity of the feature points is above a quantity threshold as valid grids; and
determining the texture distribution rating according to a quantity of the valid grids and a total quantity of the grids.

4. The method according to claim 1 , wherein the determining a texture distribution rating for the target image according to the feature points comprises:
determining, for the feature points, projection points on a diagonal line of the target image according to a predefined angle; and
determining the texture distribution rating according to distribution of the projection points along the diagonal line.

5. The method according to claim 1 , wherein the determining a texture repetition rating for the target image according to the feature point comprises:
deriving a first feature value from within a predefined range surrounding the first feature point, and deriving a second feature value from within a predefined range surrounding the second feature point;
determining whether the first feature point and the second feature point are similar feature points according to the first feature value and the second feature value; and
determining the texture repetition rating according to a quantity of the similar feature points and a total quantity of the feature points.

6. The method according to claim 5 , further comprising:
determining a spatial distance between the first feature point and the second feature point;
wherein the determining whether the first feature point and the second feature point are similar feature points according to the first feature value and the second feature value comprises:
when the first feature value and the second feature value differ by less than a first threshold and the spatial distance is above a second threshold, determining the first feature point and the second feature point to be similar feature points.

7. The method according to claim 1 , wherein the processing the target image according to a predefined processing policy to derive a contrast image comprises:
deriving a reference image by setting a predefined image to be a background image for the target image; and
injecting noise into the reference image to derive the contrast image.

8. The method according to claim 7 , wherein the injecting noise into the reference image comprises:
injecting lighting noise and/or Gaussian noise into the reference image.

9. The method according to claim 7 , wherein the determining a texture contrast rating for the target image according to the contrast image comprises:
determining a first similarity according to the reference image and the target image;
determining a second similarity according to the contrast image and the target image; and
determining the texture contrast rating for the target image according to the first similarity and the second similarity.

10. An apparatus for evaluating image tracking effectiveness, comprising:
a memory, a processor, and a computer program stored on the memory and operable on the processor,
wherein the processor, when running the computer program, is configured to:
extract feature points from a target image according to a predefined algorithm;
determine first texture information for the target image according to the feature points;
process the target image according to a predefined processing policy to derive a contrast image, and determine second texture information for the target image according to the contrast image; and
rate tracking effectiveness for the target image according to the first texture information and the second texture information;
wherein the processor is further configured to:
determine a first texture rating for the target image according to the feature points, wherein the first texture rating comprises at least one of the following texture ratings:
a texture richness rating, a texture distribution rating, and a texture repetition rating;
determine a texture contrast rating for the target image according to the contrast image;
determine a lowest rating among the first texture rating and the texture contrast rating as a rating for the target image.

11. The apparatus according to claim 10 , wherein the processor is further configured to:
derive a quantity of the feature points; and determine the texture richness rating according to the quantity.

12. The apparatus according to claim 10 , wherein the processor is further configured to:
equally partition the target image into a number of grids;
derive a quantity of the feature points in each of the grids, and determine the grids in which the quantity of the feature points is above a quantity threshold as valid grids; and
determine the texture distribution rating according to a quantity of the valid grids and a total quantity of the grids.

13. The apparatus according to claim 10 , wherein the processor is further configured to:
determine, for the feature points, projection points on a diagonal line of the target image according to a predefined angle; and
determine the texture distribution rating according to distribution of the projection points along the diagonal line.

14. The apparatus according to claim 10 , wherein the processor is further configured to:
derive a first feature value from within a predefined range surrounding the first feature point, and derive a second feature value from within a predefined range surrounding the second feature point;
determine whether the first feature point and the second feature point are similar feature points according to the first feature value and the second feature value; and
determine the text

Assignees

• Baidu online network technology beijing co ltd

Inventors

• Liu Zhaoliang
• Chen Sili
• Zhang Yongjie

Classifications

• G06V10/462

  Salient features, e.g. scale invariant feature transforms [SIFT] · CPC title

• G06V10/7515

  Shifting the patterns to accommodate for positional errors · CPC title

• G06V10/772

  Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries · CPC title

• G06V10/62

  relating to a temporal dimension, e.g. time-based feature extraction; Pattern tracking · CPC title

• G06T7/246 Primary

  using feature-based methods, e.g. the tracking of corners or segments · CPC title