IP Library Granted Patent US 12707028
Granted Patent B2
US 12707028 · App. 19/183,266 · Granted Aug 11, 2026

Artificial intelligence image analysis system and method using a preprocessing model of CCTV camera image data

Inventor: Seung Hwan Lee (Seoul, KR)
H04N7/181G06V20/52
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 12707028
App. No.
19/183,266
Granted
Aug 11, 2026
Kind
B2
Abstract

Provided is an artificial intelligence image analysis system and method using a preprocessing model of CCTV camera image data, which not only enhances the performance of the artificial intelligence network model but also improves reliability in response to changes in external environments. The system includes: a CCTV image collection unit configured to collect images captured through a plurality of CCTV cameras installed in designated zones across a plurality of regions; and an image analysis server configured to update an AI CCTV image analysis model to analyze the images by using an AI preprocessing model that is configured to receive raw images collected by the CCTV image collection unit, learn backgrounds and objects in the images based on a reference image of the captured images of the designated zones, and check status information of the CCTV camera.

Claims (37)

1 . An artificial intelligence image analysis system using a preprocessing model of CCTV camera image data, the system comprising:

a CCTV image collection unit configured to collect images captured through a plurality of CCTV cameras installed in designated zones across a plurality of regions; and

an image analysis server configured to update an AI CCTV image analysis model to analyze the images by using an AI preprocessing model,

wherein the AI preprocessing model is configured to receive raw images collected by the CCTV image collection unit, learn backgrounds and objects in the images based on a reference image of the captured images of the designated zones, and check status information of the CCTV camera, and

wherein the image analysis server includes:

an image database configured to store the images received from the CCTV image collection unit;

an AI preprocessing model unit configured to learn the backgrounds and the objects in the designated zones based on the images stored in the image database and the reference image;

a camera status determination unit configured to extract the status information of the CCTV camera from data generated by the AI preprocessing model unit;

an alarm unit configured to generate alarm information when the camera status determination unit determines that a status of the CCTV camera is abnormal;

a camera control unit configured to generate a camera control command signal for restoring the status of the CCTV camera to a status corresponding to the reference image based on the status information of the CCTV camera determined as abnormal, and transmit the camera control command signal to the CCTV image collection unit;

an optimization unit configured to update the AI CCTV image analysis model by training the AI CCTV image analysis model based on data learned by the AI preprocessing model unit; and

an analysis unit configured to analyze the images by using the updated AI CCTV image analysis model.

2 . The system of claim 1 ,

wherein the image collection unit is configured to:

transmit the captured images to the image analysis server through a video management system (VMS) or an edge device; or

transmit the captured images by the CCTV camera to the image analysis server using dual RTSP (Real-Time Streaming Protocol).

3 . The system of claim 1 ,

wherein the status information of the CCTV camera includes at least one of zoom information, angle of view information, screen obstruction information, focus information, or a combination thereof, of the image captured by the CCTV camera.

4 . An artificial intelligence image analysis method using a preprocessing model of CCTV camera image data, the method comprising:

(a) collecting, by a CCTV image collection unit, images captured through a plurality of CCTV cameras installed in designated zones across a plurality of regions; and

(b) updating, by an image analysis server, an AI CCTV image analysis model to analyze the images by using an AI preprocessing model,

wherein the AI preprocessing model is configured to receive raw images collected by the CCTV image collection unit, learn backgrounds and objects in the images based on a reference image of the captured images of the designated zones, and check status information of the CCTV camera, and

wherein the step (b) includes:

(b1) storing, by an image database, the images received from the CCTV image collection unit;

(b2) learning, by an AI preprocessing model unit, the backgrounds and the objects in the designated zones based on the images stored in the image database and the reference image;

(b3) extracting, by a camera status determination unit, status information of the CCTV camera from data generated by the AI preprocessing model unit;

(b4) generating, by an alarm unit, alarm information when the camera status determination unit determines that a status of the CCTV camera is abnormal;

(b5) generating, by a camera control unit, a camera control command signal for restoring the status of the CCTV camera to a status corresponding to the reference image based on the status information of the CCTV camera determined as abnormal, and transmitting the camera control command signal to the CCTV image collection unit;

(b6) updating, by an optimization unit, the AI CCTV image analysis model by training the AI CCTV image analysis model based on data learned by the AI preprocessing model unit; and

(b7) analyzing, by an analysis unit, the images by using the updated AI CCTV image analysis model.

5 . The method of claim 4 ,

wherein the step (a) includes:

(a1) transmitting, by the image collection unit, the captured images to the image analysis server through a video management system (VMS) or an edge device; or

(a2) transmitting, by the CCTV camera, the captured images to the image analysis server using dual RTSP (Real-Time Streaming Protocol).

6 . The system of claim 4 ,

wherein the status information of the CCTV camera includes at least one of zoom information, angle of view information, screen obstruction information, focus information, or a combination thereof, of the image captured by the CCTV camera.

7 . A computer program stored in a non-transitory computer-readable medium for executing the artificial intelligence image analysis method using a preprocessing model of CCTV camera image data according to claim 4 .