LAPTOP DETECTION

• US 20130163811A1
• Filed: 11/21/2012
• Published: 06/27/2013
• Est. Priority Date: 11/22/2011
• Status: Active Grant

• Current Assignee: Smiths Heimann GmbH (Smiths Group International Holdings Ltd.)
• Original Assignee: Smiths Heimann GmbH (Smiths Group International Holdings Ltd.)
• Inventors: Oelke, Dominik
• Granted Patent: US 9,449,242 B2
• Time in Patent Office: Days
• US Class Current: 382/103
• CPC Class Codes: G06T 2207/10116 X-ray image
• CPC Class Codes: G06T 2207/30112 Baggage; Luggage; Suitcase
• CPC Class Codes: G06T 7/0008 checking presence/absence

Abstract

Provided herein are devices, systems, and methods for the detection of objects (e.g., laptop computers, electronics, explosives, etc.) within luggage. In particular, methods are provided for the detection of laptop computers within luggage (e.g., luggage containing other metallic objects and/or electronic devices).

Claims

1. A method of detecting the presence of an object of interest in a container comprising:

determining a final size and a final probability score of a final search region; and

determining whether the object of interest exist within the container based on the final size of the final search region, final probability score of the final search region, or both.

2. The method according to claim 1, where determining a size and a final probability score of a final search region comprises the steps of:

(a) determining a search region;

(b) calculating a search region probability score;

(c) expanding the search region to a final search region;

(d) calculating a probability score for the final search region;

(e) determining the size of the final search region.

3. The method according to claim 2, where determining a search region comprises the steps of:

(i) obtaining an image of the container;

(ii) identifying a region of the image with one or more features that are characteristic of the object of interest;

(iii) applying a minimum bounding region to the image, wherein the minimum bounding region encompasses the region of the image with one or more features that are characteristic of the object of interest;

(iv) selecting a search region within the minimum bounding region.

4. The method according to claim 2, wherein said region probability score provides a likelihood that a portion of the object of interest resides within the search region, and wherein the search region probability score is based upon properties of the object of interest.

5. The method according to claim 2, where said expanding the search region to a final search region comprises the steps of:

(i) expanding the search region in one direction;

(ii) calculating a probability score for the expanded portion of the search region;

(iii) accepting or rejecting the expanded portion into the search region based on the probability score of the expanded portion;

(iv) repeating steps (i) through (iii) until expansion in each direction consecutively;

(A) meets an edge of the minimum bound region;

(B) does not result in a probability score for the expanded portion over an acceptance threshold;

(C) a combination of (A) and (B);

(V) establishing a final search region with a final size and final probability score.

6. The method of claim 3, wherein the image is obtained by dual-energy x-ray imaging.

7. The method of claim 1, wherein the object of interest comprises a laptop computer.

8. The method of claim 1, wherein the container comprises luggage, a bag, or a box.

9. A method of detecting the presence of an object of interest in a container comprising:

(a) obtaining an image of the container;

(b) identifying a region of the image with one or more features that are characteristic of the object of interest;

(c) applying a minimum bounding region to the image, wherein the minimum bounding region encompasses the region of the image with one or more features that are characteristic of the object of interest;

(d) selecting a search region within the minimum bounding region;

(e) calculating a search region probability score, wherein the search region probability score provides a likelihood that a portion of the object of interest resides within the search region, and wherein the search region probability score is based upon properties of the object of interest;

(f) expanding the search region in one direction;

(g) calculating a probability score for the expanded portion of the search region;

(h) accepting or rejecting the expanded portion into the search region based on the probability score of the expanded portion;

(i) repeating steps (f) through (h) until expansion in each direction;

(1) meets an edge of the minimum bound region;

or

(2) does not result in a probability score for the expanded portion over an acceptance threshold, thereby establishing a final search region with a final size and final probability score; and

(k) determining whether the object of interest exist within the container based on the final size and/or final probability score of the final search region.

10. The method of claim 9, the image is obtained by dual-energy x-ray imaging.

11. The method of claim 9, wherein the object of interest comprises a laptop computer.

12. The method of claim 9, wherein the container comprises luggage, a bag, or a box.

13. A method of detecting the presence of a laptop in a piece of luggage comprising:

(a) obtaining a dual energy x-ray image of the piece of luggage;

(b) analyzing the image for characteristic features of a laptop, wherein the characteristic features comprise features selected from;

materials present, number of homogeneous segments, size of homogeneous segments, density of material, and presence of circuit board;

(c) combining the characteristic features into a combined-features laptop probability score for each pixel of the image, wherein the combined features laptop probability score is proportional to the probability that a pixel contains a portion of the laptop;

(d) binarizing the image according to a threshold combined-features laptop probability score;

(e) identifying continuous regions of pixels above the threshold combined-features laptop probability score;

(f) applying one or more minimum bounding rectangles to the image, wherein each minimum bounding rectangle encompasses a single continuous region of pixels above the threshold combined-features laptop probability score;

(g) analyzing laptop properties within the minimum bounding rectangles, wherein the laptop properties comprise properties selected from;

size of minimum bounding rectangle, mean gray value of metal image, standard deviation of gray value of metal image, fraction of pixels with very high metal content, fraction of pixels with very low metal content, fraction of pixels with very low non-metal content, and aspect ratio of the rectangle;

(h) calculating a minimum bounding rectangle laptop probability score, wherein the minimum bounding rectangle laptop probability score is a composite of the laptop properties, and wherein the minimum bounding rectangle laptop probability score provides a likelihood that the minimum bounding rectangle encompasses a laptop;

(i) creating one or more subregions within a minimum bounding rectangle;

(j) calculating subregion laptop probability scores for, wherein the subregion laptop probability score is a composite of the laptop properties, and wherein the subregion laptop probability score provides a likelihood that the subregion encompasses a laptop;

(k) selecting the subregion with the highest subregion laptop probability score to be an initial search region;

(l) expanding the initial search region in one dimension;

(m) calculating a laptop probability score for the expanded portion;

(n) accepting or rejecting the expanded portion into the search region based on the probability score of the expanded portion;

(o) expanding the search region in a dimension 90°

, 180°

, or 270°

to the previous expansion;

(p) repeating steps (m) through (o) until expansion in each direction consecutively;

(1) reaches an edge of the minimum bound rectangle,

(2) does not result in an expanded portion of suitable probability score, or

(3) a combination thereof,

thereby providing a final search region of final size and final laptop probability score; and

(q) determining whether a laptop exist within the piece of luggage based on the final size and/or final probability score of the final search region.

14. A method of detecting the presence of an object of interest in a container comprising:

(a) obtaining a dual energy x-ray image of the container;

(b) identifying regions of the image with one or more features that are characteristic of the object of interest;

(c) applying a minimum bounding rectangle to the image, wherein the minimum bounding rectangle encompasses the regions of the image with one or more features that are characteristic of the object of interest;

(d) selecting one or more subregions within the minimum bounding rectangle;

(e) calculating subregion probability scores for each the subregion, wherein subregion probability score relates to the likelihood that the subregion contains a portion of the object of interest, and wherein the subregion probability score is based upon the features that are characteristic of the object of interest;

(f) selecting the subregion with the highest probability score as the search region;

(g) expanding the search region in one dimension;

(h) calculating the probability score for the expanded portion of the search region;

(i) accepting or rejecting the expanded portion into the search region based on the probability score of the expanded portion;

(j) repeating steps (g) through (i) until expansion in every direction consecutively;

(1) reaches an edge of the minimum bound rectangle,

(2) does not result in an expanded portion of suitable probability score, or

(3) a combination thereof,

thereby providing a final search region of final size and final laptop probability score; and

(k) determining whether the object of interest exist within the contained based on the final probability score of the final search region.

15. The method of claim 14, wherein determining whether the object of interest exist within the contained based on the final probability score and the final size of the final search region.