OBJECT LOCALIZATION USING VERTICAL SYMMETRY

• US 20140161308A1
• Filed: 01/31/2013
• Published: 06/12/2014
• Est. Priority Date: 12/07/2012
• Status: Active Grant

• Current Assignee: Analog Devices, Inc.
• Original Assignee: Analog Devices, Inc.
• Inventors: Krishnamoorthy, Venkatesan
• Granted Patent: US 9,076,034 B2
• Time in Patent Office: Days
• US Class Current: 382/103
• CPC Class Codes: G06T 2207/20021 Dividing image into blocks, subimages or windows
• CPC Class Codes: G06T 2207/30196 Human being; Person
• CPC Class Codes: G06T 7/68 of symmetry
• CPC Class Codes: G06V 10/50 by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
• CPC Class Codes: G06V 20/58 Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
• CPC Class Codes: G06V 40/103 Static body considered as a whole, e.g. static pedestrian or occupant recognition

Abstract

A symmetric object in an image is identified by (a) converting an edge map of the acquired image into a binary image map including binary pixel values; (b) dividing the binary image map within a scanning window into multiple bins; (c) summing binary pixel values in each bin; and (d) identifying the symmetric object based at least in part on the summed binary pixel values in the bins.

Claims

1. A method for identifying a symmetric object in an image represented as an edge map, the method comprising:

(a) converting the edge map into a binary image map comprising binary pixel values;

(b) dividing the binary image map within a scanning window into a plurality of bins;

(c) summing binary pixel values in each bin; and

(d) identifying the symmetric object based at least in part on the summed binary pixel values in the bins.

2. The method of claim 1, wherein converting the edge map into a binary image map comprises comparing each pixel value in the edge map with a predetermined threshold value.

3. The method of claim 2, wherein the threshold is determined based at least in part on pixel values in the edge map.

4. The method of claim 1, wherein identifying the symmetric object comprises comparing the summed binary pixel values with a threshold value.

5. The method of claim 4, wherein the threshold value is predetermined.

6. The method of claim 4, wherein the threshold value is dynamically calculated based on the image.

7. The method of claim 6, wherein the threshold value satisfies the equation:

TH₁=Max(TH_a,TH_b),

where TH_a is a function of a number of pixels having values above a second threshold, TH_b represents a percentage of a number of pixels in each bin, and Max is a function that returns the larger of TH_a or TH_b.

8. The method of claim 1, further comprising comparing the summed binary pixel values of two horizontal or vertical bins having locations that are symmetric in the scanning window.

9. The method of claim 1, further comprising summing the summed binary pixel values of two horizontal or vertical bins having locations that are symmetric in the scanning window.

10. The method of claim 6, wherein identifying the symmetric object comprises comparing the summed binary pixel values of the two bins with a predetermined threshold value.

11. The method of claim 1, wherein the plurality of bins are non-overlapping.

12. The method of claim 1, wherein the plurality of bins contain an identical number of binary pixels.

14. The method of claim 1, further comprising moving the scanning window to a different location within the binary image map and repeating steps (c), (d) and (e) in each location.

15. The method of claim 1, further comprising dividing each bin into a plurality of sub-bins, the symmetric object being determined based at least in part on the summed binary pixel values in the sub-bins.

16. The method of claim 1, wherein dimensions of the scanning window are determined based on a size of the object.

13. The method of 12, wherein numbers of the horizontal and vertical bins satisfy equations:

17. A system for identifying a symmetrical object in an image, the system comprising:

computer memory for storing an edge map of the image; and

a computer processor configured to;

(a) convert the edge map into a binary image map comprising binary pixel values;

(b) divide the binary image map within a scanning window into a plurality of bins;

(c) sum binary pixel values in each bin; and

(d) identify the symmetrical object based at least in part on the summed binary pixel value in the bins.

18. The system of claim 17, wherein the binary image map is stored in the computer memory.

19. The system of claim 17, wherein the processor is further configured to compute a sum of the summed binary pixel values of two horizontal or vertical bins having locations that are symmetric in the scanning window.

20. The system of claim 17, wherein the processor is a digital-signal processor.