Door Camera Depth-Based Background Blurring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional door communication systems with CCD or CMOS sensors face challenges in producing high-quality images and ensuring data protection, as they struggle to optimize image recognition under varying environmental conditions and often make irrelevant areas of the camera's field of view unrecognizable, leading to data protection issues and reduced perceptibility of relevant areas.
Innovation Solution
The door communication system employs a camera with an object or person recognition feature that blurs or masks unrecognizable image areas based on detected geometric properties and relative arrangements, focusing on the closest object or person using a distance sensor, and adjusting aperture and exposure values to ensure high blurring of out-of-focus objects, while utilizing a multifocal system or camera array for sharp imaging at different distances, and incorporating an input device for selective focus and unsharp masking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional CCD or CMOS sensors are used in door communication systems, then the system can capture images, but the image quality deteriorates under varying environmental conditions and irrelevant areas become unrecognizable, leading to data protection issues
Solution Approach 1:
The patent applies local quality by selectively blurring only the background areas while keeping the foreground subject sharp. The processing unit distinguishes between relevant (foreground) and irrelevant (background) image areas based on depth information, applying different processing qualities to each region. This resolves the contradiction by maintaining high image quality for the subject while protecting privacy by blurring irrelevant areas.
Solution Approach 2:
The patent segments the image into foreground and background regions based on depth information from the detection device. The processing unit then applies different processing operations to each segment - sharp processing for the foreground and blurring for the background. This segmentation approach allows the system to maintain reliable image quality where needed while preventing data protection violations in other areas.
2Loss of information
If the camera captures the entire field of view, then all areas are recorded, but irrelevant areas reduce the perceptibility of relevant areas and cause data protection problems
Solution Approach 1:
The patent applies local quality by selectively blurring only the background areas while keeping the foreground subject sharp. The processing unit distinguishes between relevant (foreground) and irrelevant (background) image areas based on depth information, applying different processing qualities to each region. This resolves the contradiction by maintaining high image quality for the subject while protecting privacy by blurring irrelevant areas.
Solution Approach 2:
The patent extracts and separates the foreground subject from the background using depth information. The processing unit then applies blurring specifically to the extracted background regions while maintaining sharpness for the foreground. This extraction approach removes irrelevant information that would reduce perceptibility while preventing data protection violations.
3Measurement precision
If the camera focuses on distant objects, then far-away areas are sharp, but close objects become blurred and unrecognizable
Solution Approach 1:
The patent applies dynamics by making the focus state variable rather than fixed. The detection device continuously measures depth information, and the processing unit dynamically adjusts the focus state based on which objects are detected as relevant. When close objects are detected, the system switches to close-focus mode; when distant objects are the subject, it switches to distant-focus mode. This dynamic adaptation resolves the contradiction by ensuring appropriate focus for different measurement scenarios.
Solution Approach 2:
The patent applies preliminary action by having the detection device pre-measure depth information before final image processing. The system uses this preliminary depth data to determine which objects should be in focus and which should be blurred, allowing the processing unit to prepare the appropriate focus state in advance. This preliminary measurement enables the system to resolve the focus contradiction before image generation occurs.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This solution enhances image quality and compliance with data protection by clearly recognizing the person or object closest to the camera, improving security assessments while preventing unintentional recordings of public space, thereby maintaining data protection and optimizing image clarity for property owners.
Implementation Method 1
the camera (2) has a lens with an adjustable aperture, which allows the depth of field of the generated image to be adjusted
Implementation Method 2
the camera (2) has a lens with an adjustable aperture, which allows the depth of field of the generated image to be adjusted
Implementation Method 3
the camera (2) has a lens with an adjustable aperture, which allows the depth of field of the generated image to be adjusted
Data Source
Figure 1
AI summary
The invention relates to a door communication system with a door station (1) comprising a camera (2) for generating an image of an object (3) arranged in the camera's (2) field of view, characterized in that the camera (2) further comprises a detection device (4) for detecting a geometric property of an object (3) arranged in the camera's (2) field of view and/or for detecting a relative arrangement of an object (3) arranged in the camera's (2) field of view with respect to the door station (1), wherein the camera (2) is configured to blur or mask, depending on the detected geometric property and/or the relative arrangement, those parts of the image in the camera's (2) field of view that do not depict the object. A corresponding method is further described.