Attention Object Detection in Surveillance Image Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current surveillance camera systems lack the ability to efficiently detect and track attention objects over time, making it difficult to determine the appearance time of objects and associate relevant information, which hampers their operational effectiveness.
Innovation Solution
An information processing apparatus that includes an input unit for continuous images, an attention object detection unit to identify attention objects, and a calculation unit to determine the appearance time by comparing images at different time points, along with additional units for motion detection, person object detection, and storage for associating positions with images, enabling the system to track and output relevant information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the surveillance camera system displays images and pointers without object detection and tracking, then the system structure remains simple, but the ability to detect and track attention objects over time is insufficient
Solution Approach 1:
The system performs preliminary detection of attention objects in continuous images and pre-calculates their appearance time points by comparing with previous images. This preliminary action enables the system to prepare tracking information in advance, improving detection precision without requiring complex real-time processing during playback or review operations.
Solution Approach 2:
The patent introduces an information processing apparatus as an intermediary between the surveillance camera system and the user. This intermediary automatically detects attention objects, calculates appearance time points, and associates relevant information, thereby improving detection capabilities without directly increasing the complexity of the core surveillance camera system.
2Loss of time
If the system compares first image with previous images to calculate appearance time point, then the appearance time of attention objects can be determined accurately, but the processing time and computational load increase
Solution Approach 1:
Instead of comparing the first image with all previous images, the system selectively compares with specific previous images at calculated time points. This partial action approach determines appearance time points accurately while reducing the overall computational load and processing time required for analyzing continuous image sequences.
Solution Approach 2:
The system performs image comparison and appearance time calculation as preliminary processing during image capture, rather than as post-processing. This timing strategy reduces the perceived processing time when results are needed, as the computational work has already been completed in advance.
3Reliability
If the system detects and tracks attention objects with detailed information processing, then the surveillance effectiveness is improved, but the device complexity increases
Solution Approach 1:
The patent introduces an information processing apparatus as an intermediary that handles complex detection and tracking operations. This intermediary component improves surveillance effectiveness by automatically detecting attention objects and calculating appearance time points, while isolating the complexity from the core surveillance camera system.
Solution Approach 2:
The system performs self-service by automatically detecting attention objects, tracking their movement across continuous images, and calculating appearance time points without requiring manual intervention. This automation improves surveillance reliability while the modular design keeps individual components relatively simple.
Data Source
AI summary
An information processing apparatus includes an input unit, an attention object detection unit, and a calculation unit. The input unit is configured to input a plurality of temporally continuous images taken by an image pickup apparatus. The attention object detection unit is configured to detect an attention object as an attention target from a first image which is an image taken at a first time point out of the plurality of images input. The calculation unit is configured to compare the first image with one or more second images which are one or more images taken at a time point previous to the first time point, to calculate, as a second time point, a time point when the attention object appears in the continuous plurality of images.