Image monitoring system and surveillance camera

1. A surveillance camera comprising:

an image analyzer configured to analyze an input image;

a sound source analyzer configured to analyze a sound source in the input image or a sound source that is input separately;

a determiner configured to determine whether an abnormal image exists, an abnormal sound source exists, and the abnormal image and the abnormal sound source simultaneously exist, and when the abnormal image and the abnormal sound source are obtained, according to a result of the image analysis and the sound source analysis; and

a metadata generator configured to generate metadata comprising times when the abnormal image and the abnormal sound source are obtained, respectively, types of the abnormal image and the abnormal sound source, a number of events indicated by the abnormal image and the abnormal sound source, whether the abnormal sound source is obtained from the abnormal image or an outside of the abnormal image, and whether the abnormal image and the abnormal sound source indicate a same event, based on a result of the determination of the determiner; and

an alarm information generator configured to generate alarm information about the abnormal image and the abnormal sound source based on the metadata and the result of the determination of the determiner,

wherein the abnormal image and the abnormal sound source are predefined,

wherein the determiner performs, in response to determining that the abnormal image and the abnormal sound source simultaneously exist, multiple determinations to generate the metadata which varies according to a determination result, and the determiner generates, in response to determining that the abnormal image and the abnormal sound source do not simultaneously exist, the metadata of each of the abnormal image and the abnormal sound source,

wherein the multiple determinations comprise performing a cross-check between the abnormal image and the abnormal sound source by comparing, with a threshold value, a sum of a first probability of the abnormal image being a candidate event and a second probability of the abnormal sound source being the candidate event, and

wherein if it is determined by the determiner that the abnormal image and the abnormal sound source are simultaneously obtained, the determiner further determines whether the abnormal sound source is obtained from the abnormal image or an outside of the abnormal image.

2. The surveillance camera of claim 1 , wherein if it is determined by the determiner that the abnormal image and the abnormal sound source are simultaneously obtained, the determiner further determines whether the abnormal sound source is input from the abnormal image or an outside of the abnormal image.

3. The surveillance camera of claim 2 ,

wherein if it is determined by the determiner that the abnormal image and the abnormal sound source are simultaneously obtained, and the abnormal sound source is obtained from the outside of the abnormal image, the metadata generator generates the metadata about each of the abnormal image and the abnormal sound source.

4. The surveillance camera of claim 2 , wherein if it is determined by the determiner that the abnormal image and the abnormal sound source are simultaneously obtained, and the abnormal sound source is obtained from the abnormal image, the determiner further determines whether the abnormal image and the abnormal sound source indicate the same event.

5. The surveillance camera of claim 1 , wherein the types of the abnormal sound source and the abnormal image each comprises a plurality of candidates and accuracy information about the plurality of candidates, and

wherein the determiner determines whether the abnormal image and the abnormal sound source indicate the same event with reference to the plurality of candidates corresponding to the types of the abnormal image and the abnormal sound source and the accuracy information.

6. The surveillance camera of claim 5 , wherein if it is determined by the determiner that the sum of the first probability of the abnormal image corresponding to the candidate event and the second probability of the abnormal sound source corresponding to the candidate event exceeds the threshold value, the determiner further determines that the abnormal image and the abnormal sound source indicate the candidate event.

7. The surveillance camera of claim 1 , wherein if it is determined by the determiner that the abnormal image and the abnormal sound source are simultaneously obtained, and the abnormal sound source is obtained from the outside of the abnormal image, the metadata generator generates the metadata about each of the abnormal image and the abnormal sound source.

8. The surveillance camera of claim 1 , wherein if it is determined by the determiner that the abnormal image and the abnormal sound source are simultaneously obtained, and the abnormal sound source is obtained from the abnormal image, the determiner further determines whether the abnormal image and the abnormal sound source indicate the same event, and if it is determined by the determiner that the abnormal image and the abnormal sound source indicate the same event, the metadata generator reduces the number of events indicated by the abnormal image and the abnormal sound source from the metadata.

9. The surveillance camera of claim 8 , wherein if it is determined by the determiner that the abnormal image and the abnormal sound source are simultaneously obtained, the abnormal sound source is obtained from the abnormal image, and the abnormal sound source and the abnormal image indicate different events, the metadata generator generates the metadata about each of the abnormal image and the abnormal sound source.

10. A monitoring system comprising:

the surveillance camera of claim 1 ;

an image recording apparatus configured to receive the input image, the metadata and the alarm information from the surveillance camera, store the input image, the metadata and the alarm information, and perform processing for displaying the alarm information; and

a display apparatus configured to display the input image, the metadata, and the alarm information.

11. The monitoring system of claim 10 , wherein if it is determined by the determiner that the abnormal image and the abnormal sound source are simultaneously obtained, the determiner further determines whether the abnormal sound source is input from the abnormal image or an outside of the abnormal image.

12. The monitoring system of claim 11 ,

wherein if it is determined by the determiner that the abnormal image and the abnormal sound source are simultaneously obtained, and the abnormal sound source is obtained from the outside of the abnormal image, the metadata generator generates metadata about each of the abnormal image and the abnormal sound source.

13. The monitoring system of claim 11 , wherein if it is determined by the determiner that the abnormal image and the abnormal sound source are simultaneously obtained, and the abnormal sound source is obtained from the abnormal image, the determiner further determines whether the abnormal image and the abnormal sound source indicate a same event.

14. A method of controlling a surveillance camera, the method comprising:

analyzing an input image;

analyzing a sound source in the input image or a sound source that is input separately;

determining whether an abnormal image exists, an abnormal sound source exists, and the abnormal image and the abnormal sound source simultaneously exist, and when the abnormal image and the abnormal sound source are obtained, according to a result of the image analysis and the sound source analysis;

generating metadata comprising times when the abnormal image and the abnormal sound source are obtained, respectively, types of the abnormal image and the abnormal sound source, a number of events indicated by the abnormal image and the abnormal sound source, whether the abnormal sound source is obtained from the abnormal image or an outside of the abnormal image, and whether the abnormal image and the abnormal sound source indicate a same event, based on a result of the determining; and

generating alarm information about the abnormal image and the abnormal sound source based on the metadata and the result of the determining,

wherein the abnormal image and the abnormal sound source are predefined,

wherein the generating comprises performing, in response to determining that the abnormal image and the abnormal sound source simultaneously exist, multiple determinations to generate metadata which varies according to a determination result, and generating, in response to determining that the abnormal image and the abnormal sound source do not simultaneously exist, metadata of each of the abnormal image and the abnormal sound source,

wherein the multiple determinations comprise performing a cross-check between the abnormal image and the abnormal sound source by comparing, with a threshold value, a sum of a first probability of the abnormal image being a candidate event and a second probability of the abnormal sound source being the candidate event, and

wherein method further comprises, if it is determined that the abnormal image and the abnormal sound source are simultaneously obtained, determining whether the abnormal sound source is obtained from the abnormal image or an outside of the abnormal image.

15. The method of claim 14 , wherein the determining comprises determining whether the abnormal sound source is obtained from the abnormal image or an outside of the abnormal image.

16. The method of claim 15 , wherein the determining comprises determining whether the abnormal image and the abnormal sound source indicate the same event.

17. The method of claim 14 , wherein the types of the abnormal sound source and the abnormal image each comprises a plurality of candidates and accuracy information about the plurality of candidates, and

wherein the determining comprises determining whether the abnormal image and the abnormal sound source indicate the same event with reference to the plurality of candidates corresponding to the types of the abnormal image and the abnormal sound source and the accuracy information.