Doorbell Camera Dual-Region Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional doorbell systems often experience detection errors due to motion detection based on illuminance changes, leading to incorrect image capture and blank video calls when users attempt to communicate with visitors.
Innovation Solution
A doorbell apparatus with a camera that captures images in two regions, using separate image processing units for object detection and identification, reducing erroneous detection by applying different algorithms to each area and storing identified images for accurate object recognition and alarm activation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If motion detection technique is used to generate signal for capturing image, then image capture is triggered, but detection errors occur when illuminance changes are mistaken for object motion
Solution Approach 1:
The image is divided into multiple regions (first region, second region, third region) at different distances from the camera. Different image processing algorithms are applied to each region: the first image processing unit detects objects in the first region, the second image processing unit identifies objects in the second region, and the third image processing unit detects objects in the third region. This segmentation resolves the contradiction by allowing specialized processing for each distance zone, improving both reliability and measurement precision simultaneously.
2Measurement precision
If a single image processing unit is used, then device complexity is reduced, but detection accuracy and object identification capability deteriorate
Solution Approach 1:
The image processing system is segmented into multiple specialized units (first, second, and third image processing units) that operate on different image regions. Each unit is optimized for its specific function: detection in distant regions and identification in closer regions. This segmentation improves measurement precision while managing device complexity through functional specialization rather than a single complex unit handling all tasks.
Solution Approach 2:
Different image processing algorithms and processing strategies are applied to different regions of the image based on distance from the camera. The first image processing unit applies algorithms suitable for distant object detection, while the second image processing unit applies algorithms optimized for identifying objects in the second region. This local quality approach ensures each region receives appropriate processing, improving overall accuracy without requiring a single overly complex system.
3Loss of time
If object detection is performed at far distance, then early detection is achieved, but identification accuracy decreases
Solution Approach 1:
The first image processing unit performs preliminary object detection in the first region (far distance) to identify potential objects of interest early. When an object is detected, the system then activates the second image processing unit to perform detailed identification in the second region (closer distance). This preliminary action approach reduces time loss by detecting objects early while maintaining identification precision through subsequent closer examination.
Solution Approach 2:
The detection and identification processes are segmented into two distinct stages performed by different image processing units. The first unit handles distant detection to minimize time loss, while the second unit handles closer identification to ensure precision. This segmentation allows the system to optimize for both speed and accuracy at different stages of the object recognition process.
Data Source
AI summary
A doorbell apparatus includes a camera to acquire a first image of an object in a first region and a second image of the object in a second region disposed closer to the camera than the first region; an image processing unit to detect the object in the first image and to identify the object in the second image, the image processing unit having a first image processing unit that detects and saves the first images and a second image processing unit that identifies and saves objects in the second images; a storage device to store the second image corresponding to the identified object; and an alarming unit to activate an alarm signal depending on whether the identified object corresponds to a predetermined object.