Dual-Sensor Image Capture for Object Detection Bandwidth Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image capture systems face significant bandwidth, cost, and power consumption burdens due to continuous high-resolution data uploads for cloud-based image analysis, and often struggle to capture images with sufficient resolution for recognizing specific objects like faces.
Innovation Solution
An image capture device employs a wide-angle sensor to capture a low-resolution initial image, which is analyzed to identify regions of interest using object detection models. These coordinates are used to position a second sensor to capture high-resolution images of the detected objects, which are then uploaded for further processing, thereby reducing bandwidth and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution images are continuously captured and uploaded for cloud-based analysis, then object recognition accuracy is improved, but bandwidth consumption and power usage increase significantly
Solution Approach 1:
The patent divides the image capture process into two stages: first capturing a low-resolution wide-field image to identify regions of interest, then capturing high-resolution images only of those specific regions. This segmentation approach allows the system to maintain object recognition accuracy while significantly reducing the total amount of high-resolution data that needs to be processed and transmitted, thereby reducing power consumption.
Solution Approach 2:
Instead of continuously capturing and uploading all high-resolution images, the system performs partial action by only capturing high-resolution images of regions that contain objects of interest. This selective approach reduces the overall data processing load and power consumption while maintaining sufficient object recognition accuracy for the identified regions.
2Measurement precision
If high-resolution images are continuously uploaded to the cloud, then object detection quality is improved, but bandwidth consumption increases significantly
Solution Approach 1:
The patent extracts and processes only the relevant portions of the visual scene by first identifying regions of interest in low-resolution images, then capturing high-resolution images only of those extracted regions. This extraction approach reduces the quantity of data that needs to be transmitted over the network while maintaining object detection quality for the regions that matter most.
Solution Approach 2:
The system performs partial action by uploading only the necessary high-resolution image data (regions of interest) to the cloud rather than continuously uploading all high-resolution images. This reduces bandwidth consumption while maintaining sufficient object detection quality for the identified regions.
3Area of stationary object
If a wide-angle sensor is used to capture a wide field of view, then coverage area is improved, but image resolution for specific objects decreases
Solution Approach 1:
The patent segments the imaging task into two parts: a wide-angle sensor captures the entire field of view to identify regions of interest, then a second sensor captures high-resolution images of those specific regions. This segmentation allows the system to maintain both wide coverage and high resolution for specific objects.
Solution Approach 2:
The patent transitions from a two-dimensional trade-off (wide angle vs. resolution in a single image) to a three-dimensional solution by adding a temporal dimension: first capturing a wide low-resolution image, then capturing a high-resolution image of the region of interest. This dimensional change allows both wide coverage and high resolution to be achieved at different times in the imaging sequence.
4Measurement precision
If continuous high-resolution image streams are transmitted to the cloud, then comprehensive object analysis is improved, but processing cost increases significantly
Solution Approach 1:
The patent extracts only the essential data (regions containing objects of interest) from the full visual scene and transmits only those extracted regions to the cloud for analysis. This extraction approach reduces data transmission volume while maintaining comprehensive object analysis for the identified regions.
Solution Approach 2:
The system performs partial action by transmitting only the necessary image data (high-resolution regions of interest) to the cloud rather than continuously transmitting all high-resolution images. This reduces processing cost while maintaining sufficient object analysis comprehensiveness for the identified regions.
Data Source
AI summary
A method and system for capturing an image of a portion of an environment including an object of interest and uploading of the image to a service for further image analysis. The system captures a first image of an environment or scene using a first sensor of an image capture device. A region of interest is detected within the first image using an object detection model. A set of coordinates corresponding to the portion of the first image is identified and used to position a second sensor. The second sensor captures a second image including the targeted region of interest, wherein the second image has a higher resolution than the first image. The second image is uploaded to an object detection service for the further image analysis.