IP Library Granted Patent US 9,031,285
Granted Patent B2
US 9,031,285 · App. 13/864,551 · Granted May 12, 2015

Detection of floating objects in maritime video using a mobile camera

Inventors: Hieu Tat Nguyen (Cypress, CA); Prakash Ramu (Sunnyvale, CA); Hai Wei (Los Angeles, CA); Xiaoqing Liu (El Cerrito, CA); Jacob Yadegar (Santa Monica, CA)
Assignee: UtopiaCompression Corporation
G06K9/0063
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 9,031,285
App. No.
13/864,551
Granted
May 12, 2015
Kind
B2
Abstract

A method and system for detecting floating objects in maritime video is disclosed. The horizon is detected within the video. Modeling of the sky and water is performed on the video. Objects are detected that are not water and sky within the video.

Claims (45)

1. In a camera system, a method of detecting objects in a maritime video stream comprising the steps of:

acquiring with a video camera, a maritime video stream;

providing said maritime video stream to a central processing unit that is integrated with said video camera, said central processing unit running a software application which is operative for

receiving said maritime video stream,

detecting a horizon within said maritime video stream to distinguish a water region from a sky region;

modeling the water region and the sky region within the maritime video, wherein modeling of the sky region and the water region uses two separate regressions, and

detecting objects that are not water and not sky within the maritime video stream.

2. The method of claim 1 , wherein the step of detecting the horizon is performed prior to the step of modeling the water region and the sky region.

3. The method of claim 1 , further comprising segmenting the maritime video stream into a water region, a sky region and a land region.

4. The method of claim 1 , wherein the step of modeling the water region and the sky region, comprises performing a global regression of spatial image coordinates.

5. The method of claim 4 , wherein the step of modeling the water region and the sky region is based on weighted least squares.

6. The method of claim 4 , wherein the step of performing a global regression is completed using an estimation method.

7. The method of claim 6 wherein said estimation method is a weighted least squares method.

8. The method of claim 1 , further comprising correcting the horizon based on the modeling of the water region and the sky region.

9. The method of claim 1 , wherein the step of detecting objects comprises identifying an object as a flying target if it is surrounded by sky.

10. The method of claim 1 , wherein the step of detecting objects comprises identifying an object as a maritime target if it is surrounded by water.

11. The method of claim 1 , wherein modeling the sky region and the water region is performed for each color channel.

12. The method of claim 1 , wherein the modeling of the sky and water regions further comprises determining residue for each pixel within the maritime video stream and whether the residue of the pixel exceeds a predetermined threshold.

13. The method of claim 12 , further comprising segmenting the pixel where the residue of the pixel exceeds the predetermined threshold as land and object regions.

14. The method of claim 1 , further comprising analyzing the residue of the water model to correct the horizon.

15. The method of claim 1 , further comprising filtering the water and sky models.

16. The method of claim 15 , wherein filtering the water and sky models is accomplished by applying an open-close smoothing filter to the water and sky models.

17. In a camera system, a method of detecting objects in a maritime video stream comprising the steps of:

acquiring with a video camera, a maritime video stream;

providing said maritime video stream to a central processing unit that is integrated with said video camera, said central processing unit running a software application which is operative for

receiving said maritime video stream,

detecting a horizon within said maritime video stream to distinguish a water region from a sky region;

modeling the water region and the sky region within the maritime video; and

detecting objects that are not water and not sky within the maritime video stream,

wherein the step of detecting objects comprises marking a region as a land area if a significant portion of said region is above the horizon and the region is connected to a left edge or a right edge of the video.

18. The method of claim 17 , wherein the step of detecting objects comprises, marking portions of the land region that are above the horizon as land and marking portions of the land region that are below the horizon as maritime targets.

19. In a camera system, a method of detecting objects in a maritime video stream comprising the steps of:

acquiring with a video camera, a maritime video stream;

providing said maritime video stream to a central processing unit that is integrated with said video camera, said central processing unit running a software application which is operative for

receiving said maritime video stream,

detecting a horizon within the maritime video stream to distinguish a water region from a sky region,

modeling the water region and the sky region wherein the step of modeling comprises performing a global regression of spatial image coordinates based on a weighted least squares method, and further wherein modeling of the sky region and the water region uses two separate regressions, and

detecting objects that are not water and not sky within said maritime video stream, wherein the step of detecting objects further comprises segmenting potential foreground objects using hysteresis thresholding.

20. In a camera system, a method of detecting objects in a maritime video stream comprising the steps of:

acquiring with a video camera, a maritime video stream,

providing said maritime video stream to a central processing unit that is integrated with said video camera, said central processing unit running a software application which is operative for

receiving said maritime video stream,

detecting a horizon within the maritime video stream to distinguish a water region from a sky region,

modeling the water region and the sky region wherein the step of modeling comprises performing a global regression of spatial image coordinates using an estimation method, and further wherein modeling of the sky region and the water region uses two separate regressions, and

detecting objects that are not water and not sky within said maritime video stream, wherein the step of detecting objects further comprises segmenting potential foreground objects using a thresholding method.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 17, 2013
From: WEI, HAI; NGUYEN, HIEU TAT; YADEGAR, JACOB; RAMU, PRAKASH; LIU, XIAOQING
To: UTOPIACOMPRESSION CORPORATION
Reel/Frame 030233/0920 →
Continuity (1)
Related Publication 20140314270A1 · Oct 23, 2014