Image processing method and apparatus for object detection

Patent metadata

Field	Value
Publication number	US-11908117-B2
Application number	US-202117227704-A
Country	US
Kind code	B2
Filing date	Apr 12, 2021
Priority date	Mar 27, 2017
Publication date	Feb 20, 2024
Grant date	Feb 20, 2024

Abstract

An image processing method implemented by a processor includes receiving an image, acquiring a target image that includes an object from the image, calculating an evaluation score by evaluating a quality of the target image, and detecting the object from the target image based on the evaluation score.

First claim

What is claimed is:

1. A processor implemented image processing method, comprising:
receiving an image;
acquiring a target image that includes an object from the image by normalizing a target area including the object based on an information associated with a feature of the object;
calculating an evaluation score by evaluating a quality of the target image; and
detecting the object from the target image based on the evaluation score,
wherein the acquiring of the target image comprises:
performing an approximation of the target area by enlarging or reducing the target area;
performing a scale normalization on the approximated target area based on the information; and
acquiring the target image by performing an illumination normalization on the scaled normalized target area.

2. The method of claim 1 , wherein the acquiring of the target image comprises:
acquiring the target area including the object from the image; and
extracting the information associated with the feature of the object from the target area.

3. The method of claim 2 , wherein the acquiring of the target area comprises:
detecting the object included in the image; and
acquiring the target area by detecting the feature of the object from the object.

4. The method of claim 2 , wherein the information comprises any one or any combination of two or more of a contour of the feature of the object, a relative location of the feature of the object on the target area, and an area of the feature of the object on the target area.

5. The method of claim 1 , wherein the performing of the scale normalization comprises:
correcting a size and an angle of the feature of the object with respect to the approximated target area.

6. The method of claim 1 , wherein the detecting of the object from the target image comprises:
comparing the evaluation score and a detection reference score; and
determining whether to perform a detection of the object based on a result of the comparing.

7. The method of claim 6 , wherein the determining whether to perform the detection comprises:
suspending the detection of the object in response to the evaluation score being less than the detection reference score based on the comparison result; and
performing the detection of the object in response to the evaluation score being greater than or equal to the detection reference score based on the comparison result.

8. The method of claim 7 , wherein the determining whether to perform the detection further comprises:
determining whether to perform a detection of an object included in a subsequent frame based on an evaluation score of the object included in the subsequent frame, in response to the evaluation score being less than the detection reference score based on the comparison result.

9. The method of claim 1 , wherein the detecting comprises detecting the object using an AdaBoost algorithm.

10. An image processing apparatus comprising:
a receiver configured to receive an image; and
a controller configured to detect an object included in the image by evaluating a quality of the image, the controller comprising:
a target image acquirer configured to acquire a target image that includes the object from the image by normalizing a target area including the object based on an information associated with a feature of the object;
a quality evaluator configured to calculate an evaluation score by evaluating a quality of the target image; and
an object detector configured to detect the object from the target image based on the evaluation score,
wherein the target image acquirer comprises a normalizer configured to:
perform an approximation by enlarging or reducing the target area,
perform a scale normalization on the approximated target area based on the information, and
acquire the target image by performing an illumination normalization on the scaled normalized target area.

11. The image processing apparatus of claim 10 , wherein the target image acquirer comprises:
a target area acquirer configured to acquire a target area including the object from the image; and
an information extractor configured to extract information associated with a feature of the object from the target area.

12. The image processing apparatus of claim 11 , wherein the target area acquirer is configured to detect the object included in the image, and to acquire the target area by detecting the feature of the object from the object.

13. The image processing apparatus of claim 11 , wherein the information comprises any one or any combination of two or more of a contour of the feature of the object, a relative location of the feature of the object on the target area, and an area of the feature of the object on the target area.

14. The image processing apparatus of claim 10 , wherein the normalizer is configured to correct a size and an angle of the feature of the object with respect to the approximated target area.

15. The image processing apparatus of claim 10 , wherein the object detector is configured to compare the evaluation score and a detection reference score, and to determine whether to perform a detection of the object based on a result of the comparing.

16. The image processing apparatus of claim 15 , wherein the object detector is configured to suspend the detection of the object in response to the evaluation score being less than the detection reference score based on the comparison result, and to perform the detection of the object in response to the evaluation score being greater than or equal to the detection reference score based on the comparison result.

17. The image processing apparatus of claim 16 , wherein the object detector is configured to determine whether to perform a detection of an object included in a subsequent frame based on an evaluation score of the object included in the subsequent frame, in response to the evaluation score being less than the detection reference score based on the comparison result.

18. The image processing apparatus of claim 10 , wherein the object detector is configured to detect the object using an AdaBoost algorithm.

19. An image processing method implemented by a processor, comprising:
receiving an image in a plurality of frames;
acquiring a target image that includes an object from the image in a first frame of the plurality of frames by normalizing a target area including the object based on an information associated with a feature of the object;
calculating an evaluation score by evaluating a quality of the target image in the first frame; and
suspending a detection of the object from the target image in the first frame in response to the evaluation score being less than a detection reference score and performing the detection of the object in response to the evaluation score being greater than or equal to the detection reference score,
wherein the acquiring of the target image comprises:
performing an approximation of the target area by enlarging or reducing the target area;
performing a scale normalization on the approximated target area based on the information; and
acquiring the target image by performing an illumination normalization on the scaled normalized target area.

20. The method of claim 19 , wherein subsequent frames of the plurality of frames are evaluated based on the evaluation score of the object included in the first frame.

Assignees

• Samsung Electronics Co Ltd

Inventors

• Feng Hao
• Han Jae-Joon
• Choi Changkyu
• Zhang Chao
• Xu Jingtao
• Shan Yanhu
• An Yaozu

Classifications

• G06T7/0002 Primary

  Inspection of images, e.g. flaw detection · CPC title

• G06F18/2148

  characterised by the process organisation or structure, e.g. boosting cascade · CPC title

• G06T3/4007

  based on interpolation, e.g. bilinear interpolation (image demosaicing G06T3/4015; edge-driven or edge-based scaling G06T3/403) · CPC title

• G06T7/73

  using feature-based methods · CPC title

• G06V10/255

  Detecting or recognising potential candidate objects based on visual cues, e.g. shapes · CPC title