IMAGE FOREGROUND DETECTION APPARATUS AND METHOD AND ELECTRONIC DEVICE

• US 20190311492A1
• Filed: 06/21/2019
• Published: 10/10/2019
• Est. Priority Date: 01/23/2017
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: ZHANG, Nan, Wang, Qi
• Granted Patent: US 11,107,237 B2
• Time in Patent Office: Days
• CPC Class Codes: G06T 2207/20076 Probabilistic image processing
• CPC Class Codes: G06T 2207/30168 Image quality inspection
• CPC Class Codes: G06T 7/0002 Inspection of images, e.g. flaw detection
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/136 involving thresholding
• CPC Class Codes: G06T 7/194 involving foreground-background segmentation
• CPC Class Codes: G06T 7/20 Analysis of motion motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51
• CPC Class Codes: G06T 7/73 using feature-based methods

Abstract

An image foreground detection apparatus and method and an electronic device. In the apparatus and method, in updating the background model, a pixel value of a sample value of which the pixel value difference is maximum is replaced with a predetermined probability, which may efficiently improve accuracy of the background model, and relatively large number of integral foreground image blocks which are accurate may be quickly obtained.

Claims

1. An image foreground detection apparatus, comprising:

a first detector configured to perform foreground detection on each pixel of an input image, the first detector comprising;

a first calculator configured to calculate a first difference between a pixel value of the pixel and sample values in a background model corresponding to a position where the pixel is located;

a first updating unit configured to, when the number of sample values in the background model of the first difference less than or equal to a first threshold is greater than or equal to a second threshold, replace a pixel value of a sample value in the background model of which the first difference is maximum with the pixel value of the pixel by a predetermined probability, to update the background model of the position where the pixel is located; and

a first determining unit configured to, when the number of sample values in the background model of the first difference less than or equal to the first threshold is less than the second threshold, determine the pixel as a foreground pixel.

2. The apparatus according to claim 1, wherein the first detector further comprises:

a second calculator configured to, when the number of sample values in the background model of the first difference less than or equal to the first threshold is greater than or equal to the second threshold, calculate a second difference between the pixel value of the pixel and pixel values of neighboring pixels of the pixel, to determine a neighboring pixel of which the second difference is maximum; and

a second updating unit configured to calculate a third difference between the pixel value of the neighboring pixel of which the second difference is maximum and pixel values of sample values in the background model corresponding to a position where the neighboring pixel is located, and replace a pixel value of a sample value in the background model of which the third difference is maximum with the pixel value of the neighboring pixel of which the second difference is maximum by a predetermined probability, to update the background model of the position where the neighboring pixel is located.

3. The apparatus according to claim 1, wherein the first detector further comprises:

an acquiring unit configured to obtain pixel values of a first number of surrounding pixels of the pixel; and

a second determining unit configured to remove at least one surrounding pixel of which pixel value is maximum and at least one surrounding pixel of which pixel value is minimum to obtain pixel values of a second number of surrounding pixels, and determine the pixel values of the second number of surrounding pixels as being initial values of the sample values in the background model of the position where the pixel is located.

4. The apparatus according to claim 1, wherein the first detector further comprises:

a third updating unit configured to update the first threshold at a predetermined period according to clarity of the input image.

5. The apparatus according to claim 4, wherein the third updating unit updates the first threshold according to a formula as below:

6. The apparatus according to claim 1, wherein the input image comprises continuous a plurality of frames of input images, and the apparatus further comprises:

a second detector configured to perform ghost detection on each of foreground pixels that are determined as being foreground pixels in the input images; and

wherein, in performing ghost detection on each foreground pixel, when the number of times that pixels at a position where the foreground pixel is located are continuously detected as foreground pixels in the continuous a plurality of frames of input images is greater than a third threshold, or when the number of times that a difference between a pixel value of the pixel at a position where the foreground pixel is located in the continuous a plurality of frames of input images and a pixel value of a pixel at a corresponding position in a preceding frame of input image is less than a fourth threshold is greater than a fifth threshold, the foreground pixel is determined as being a ghost pixel, or else the foreground pixel is determined as being not a ghost pixel.

7. The apparatus according to claim 6, wherein the first updating unit is further configured to update a background model of a position where the foreground pixel that is determined as being a ghost pixel is located.

8. An electronic device, comprising the apparatus as claimed in claim 1.

9. An image foreground detection method, comprising:

performing foreground detection on each pixel of an input image the performing foreground detection on each pixel comprising;

calculating a first difference between a pixel value of the pixel and sample values in a background model corresponding to a position where the pixel is located;

when the number of sample values in the background model of the first difference less than or equal to a first threshold is greater than or equal to a second threshold, replacing a pixel value of a sample value in the background model of which the first difference is maximum with the pixel value of the pixel by a predetermined probability, to update the background model of the position where the pixel is located; and

when the number of sample values in the background model of the first difference less than or equal to the first threshold is less than the second threshold, determining the pixel as a foreground pixel.

10. The method according to claim 9, wherein the performing foreground detection on each pixel further comprises:

when the number of sample values in the background model of the first difference less than or equal to the first threshold is greater than or equal to the second threshold, calculating a second difference between the pixel value of the pixel and pixel values of neighboring pixels of the pixel, to determine a neighboring pixel of which the second difference is maximum; and

calculating a third difference between the pixel value of the neighboring pixel of which the second difference is maximum and pixel values of sample values in the background model corresponding to a position where the neighboring pixel is located, and replacing a pixel value of a sample value in the background model of which the third difference is maximum with the pixel value of the neighboring pixel of which the second difference is maximum by a predetermined probability, to update the background model of the position where the neighboring pixel is located.

11. The method according to claim 9, wherein the performing foreground detection on each pixel further comprises:

obtaining pixel values of a first number of surrounding pixels of the pixel; and

removing at least one surrounding pixel of which pixel value is maximum and at least one surrounding pixel of which pixel value is minimum to obtain pixel values of a second number of surrounding pixels, and determining the pixel values of the second number of surrounding pixels as being initial values of the sample values in

15. The method according to claim 14, wherein the method further comprises:

updating a background model of a position where the foreground pixel that is determined as being a ghost pixel is located.

12. The method according to claim 9, wherein the performing foreground detection on each pixel further comprises:

updating the first threshold at a predetermined period according to clarity of the input image.

13. The method according to claim 12, wherein the first threshold is updated according to a formula as below:

14. The method according to claim 9, wherein the input image comprises continuous a plurality of frames of input images, and the method further comprises:

performing ghost detection on each of foreground pixels that are determined as being foreground pixels in the input images; and

wherein, the performing ghost detection on each of the foreground pixels comprises;

when the number of times that pixels at a position where the foreground pixel is located are continuously detected as foreground pixels in the continuous a plurality of frames of input images is greater than a third threshold, or when the number of times that a difference between a pixel value of the pixel at a position where the foreground pixel is located in the continuous a plurality of frames of input images and a pixel value of pixel at a corresponding position in a preceding frame of input image is less than a fourth threshold is greater than a fifth threshold, determining the foreground pixel as being a ghost pixel, or else determining the foreground pixel as being not a ghost pixel.

16. A detection apparatus comprising:

a memory; and

a processor coupled to the memory and configured to perform a process, including;

receiving an input image containing a plurality of pixels;

calculating, for each pixel, a first difference between a pixel value of the pixel and sample values in a background model corresponding to a position where the pixel is located;

when the number of sample values in the background model of the first difference is less than or equal to a first threshold which is greater than or equal to a second threshold, replacing a pixel value of a sample value in the background model of which the first difference is maximum with the pixel value of the pixel by a predetermined probability, to update the background model of the positon where the pixel is located; and

when the number of sample values in the background model of the first difference is less than or equal to the first threshold which is less than the second threshold, determining the pixel as a foreground pixel.

the background model of the position where the pixel is located.