Image pickup system for locking object in image based on IPC

Abstract

The invention discloses a camera system for locking an object in an image based on an IPC, and relates to the field of monitoring, the camera system comprises a lens, a processing board and a support, the processing board comprises a sensing module, an identification module, a tracking module, a taking module and a decision module, the sensing module is used for sensing a moving target and detecting vehicles within 200 meters, pedestrians within 100 meters and faces within 100 meters; the recognition module is used for detecting an object, and a matching object is any one of a vehicle, a face and a behavior; the tracking module is used for locking specific personnel, vehicles and non-motor vehicles, triggering tracking, tripping lines, area invasion, article leaving, article taking, wandering, area entering, area leaving and rapid movement; the decision module is used for locking a risk target and predicting a target behavior, and the selected face is used as the input of the risk target.

Description

technical field [0001] The invention relates to the technical field of cameras, in particular to a camera system for locking objects in images based on IPC. Background technique [0002] AI technology has become the driving force for security development, and active monitoring has gradually replaced passive monitoring. [0003] 1. In the analog monitoring stage (1979-1983), the security industry gradually formed a product matrix with analog cameras + video matrix + tape recorders. [0004] 2. In the digital monitoring stage (1984-1996), DVR products using digital recording technology replaced tape video recorders and produced breakthroughs in various functions. [0005] 3. In the network monitoring stage (1997-2008), the video monitoring system became more complicated, and a product matrix of network camera + NVR + software system appeared. [0006] 4. In the high-definition monitoring stage (2009-2012), the security system has initially expanded into a platform applicatio...

AI Technical Summary

Problems solved by technology

[0005] 3. In the network monitoring stage (1997-2008), the video monitoring system was more complicated, and a product matrix of network camera + NVR + software system appeared

[0011] At present, the application of artificial intelligence is subject to the hit of the algorithm, but the biggest obstacle to the accuracy of the algorithm is the target pixel density and image clarity. Now almost all front-end cameras are not designed for artificial intelligence, so this problem cannot be solved. To solve this problem Hardware and imaging issues must be resolved

Images