Method and system for detection of foreign objects in maritime environments

• US 10,121,078 B2
• Filed: 05/13/2013
• Issued: 11/06/2018
• Est. Priority Date: 05/15/2012
• Status: Active Grant

• Current Assignee: Eshel Aviv, Ltd.
• Original Assignee: Eshel Aviv, Ltd.
• Inventors: Guissin, Rami, Lavi, Eitan, Stolovich, Akiva, Bason, Sharon
• Primary Examiner(s): Moyer, Andrew
• Assistant Examiner(s): Rosario, Dennis
• Application Number:
  US14/401,180
• Publication Number:
  US 20150104064A1
• Time in Patent Office: 2,003 Days
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/10048 Infrared image
• CPC Class Codes: G06T 2207/20016 Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
• CPC Class Codes: G06T 2207/30232 Surveillance
• CPC Class Codes: G06T 7/246 using feature-based methods, e.g. the tracking of corners or segments
• CPC Class Codes: G06T 7/32 using correlation-based methods
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69
• CPC Class Codes: G08G 3/02 Anti-collision systems

Abstract

The present invention provides techniques for detecting foreign objects in a region of interest in maritime environments. Image data indicative of a sequence of successively acquired images of the region of interest, is analyzed to determine candidate points of interest, and data indicative of said points is processed to identify candidate points that are adjacently accumulated in different locations in images data. Grouping data may be then generated based on the identified accumulations of candidate points indicative of a group of said candidate points. The grouping data is processed to identify spatio-temporal correlation between the points in the group and determine a corresponding track function, thereby enabling detection of a presence of a foreign object in the image data.

Claims

1. A method for detecting foreign objects in a region of interest in a maritime environment, the method comprising:

receiving image data indicative of a sequence of successively acquired images of the region of interest,

analyzing each image in said sequence of images to determine candidate image pixel points of interest

in the generated data indicative of the interpolated image data;

processing the generated data indicative of the candidate image pixel points in the image data

identifying the candidate image pixel points having adjacently located coordinates in two or more of the consecutive images, and

generating grouping data indicative of a plurality of groups of said adjacently located candidate image pixel points; and

processing the grouping data,

said processing the grouping data comprising

identifying spatio-temporal correlation between the adjacently located image pixel points in each group of candidate points for determining for at least some of said groups of candidate points micro tracks formed by the locations of the grouped points in the two or more consecutive images and being indicative of formation of short term tracks in the received images, and

generating micro tracks data indicative thereof,

wherein said analyzing of the sequence of images comprises;

applying to each image under analysis of the sequence of images one or more decimation functions and generating respective decimated image data;

computing a signal to clutter ratio for each pixel in said image under analysis and for each pixel in said decimated image data;

processing the generated signal to clutter ratios and

generating therefrom one or more respective binary image data

for said image under analysis, and

for said respective decimated image data according to an adaptive thresholding criterion; and

interpolating binary image data associated with said decimated image data, and generating data indicative of interpolated image data.

2. The method of claim 1, wherein said generation of the data indicative of the interpolated image data further comprises applying an OR logic function to the binary image data obtained for the analyzed image and for the interpolated binary images associated with decimated image data.

3. The method of claim 1, wherein said generation of the decimated image data comprises applying to each image two or more of the decimation functions having different decimation factors, said decimated image data thereby comprising two or more respective decimated image data pieces.

4. The method of claim 1, comprising carrying out the following operations:

applying spatial blobbing function to the data indicative of the interpolated image data thereby obtaining segmented blob maps comprising blob data associated with points appearing in the image;

filtering the segmented blob maps and generating respective spatiotemporal geometric centers for each blob in said maps, said geometric centers corresponding to the candidate points; and

generating binary images indicative of said spatiotemporal geometric centers.

5. The method of claim 4, wherein said operations are applied to the binary image data associated with the analyzed image and with the interpolated image data resulting from said interpolation of the binary images associated with said decimated image data.

6. The method of claim 4, wherein said operations are applied to the data resulting from the application of the OR logic function.

7. The method of claim 4, comprising generating the micro tracks by carrying out the following operations:

identifying the groups of adjacently located spatiotemporal geometric centers;

applying a smoothing function to said groups of adjacently located spatiotemporal geometric centers and morphological operators to generate respective blobs; and

mapping each blob to an oriented ellipse and determining slope and length of an elongated axis thereof and generating the micro track data accordingly.

8. The method of claim 7, comprising correlating the micro track data obtained for the different blobs to determine the continuous paths and generating the macro track data indicative thereof.

9. A computerized system for use in detecting foreign objects in a region of interest in maritime environments, the system comprising:

data input utility for receiving image data of two or more images successively acquired from said region of interest, said images having a cluttered background;

one or more processors and memory configured and operable to receive and analyze said image data using one or more modules stored in said memory, comprising;

a plot detection module for identifying one or more groups of candidate image pixel points adjacently located in regions in the interpolated binary images of said two or more images and generating data indicative thereof; and

a track detection module comprising;

a micro track detection unit configured and operable to

determine for at least some of said groups of candidate points micro tracks

formed by the locations of the grouped points in the two or more successively acquired images and

being indicative of formation of short term tracks in the received images, and

generate micro tracks data indicative thereof, and

a macro track detection unit configured and operable to

determine correlations between at least some of the micro tracks for

determining from the locations of said one or more groups of candidate image pixel points in the images

continuous paths of foreign objects by

combining the at least some of the micro tracks to thereby overcome intermittent obstruction of the foreign objects due to significant variability of the maritime environment and

enable generating track data accordingly;

wherein the plot detection module comprises;

a decimation utility for

decimating the received image data and

generating respective decimated image data;

a spatial filter utility for

determining a signal to clutter ratio for each pixel in the image data under analysis and for each pixel in the decimated image data, and

generating pixel data indicative thereof;

a threshold utility for

generating binary images for said image data under analysis and said decimated image data based on said pixel data and an adaptive threshold criterion; and

an interpolation utility for interpolating binary images associated with said decimated image data.

10. The system of claim 9, wherein the plot detection module comprises OR function utility for applying an OR logic function to the binary image data associated with the analyzed image data and the interpolated binary images, and generating data indicative thereof.

11. The system of claim 9, wherein the plot detection module comprises:

a blob analysis utility configured and operable for processing the binary image data, or the data obtained from the OR function utility, and generating blob maps comprising blob data associated with image pixel points in the received images;

an adaptive temporal filtering summing utility configured and operable for filtering the blob maps and generating respective spatiotemporal geometric centers for each blob in said maps; and

a plot identification utility configured and operable for identifying the candidate image pixel points based on the generated geometric centers.