Digital Depth Generation from Single Image Capture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional cameras face challenges in generating large-sized 2D images and stereoscopic 3D images suitable for human observation in real time, especially when dealing with multiple cameras or image matrices, due to slow and power-wasting voice coil motors, and difficulties in processing complex image inputs and maintaining focusing planes.
Innovation Solution
An image processing device with multiple image sensors and image matrices that uses digital image processing techniques to generate large-sized 2D and stereoscopic 3D images in real time, incorporating an image sensor array, pre-processing unit, depth information generator, and focusing unit to synchronize and enhance image focus and resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a voice coil motor (VCM) is used to trigger the lens to move forward and backward for depth information generation, then the lens focusing function is achieved, but the system consumes more power, moves slowly, and generates noise
Solution Approach 1:
The patent replaces the mechanical voice coil motor (VCM) system with a digital image processing system. Instead of physically moving the lens to capture images at different focal planes, the system captures a single image and uses computational algorithms to generate depth information and refocused images. This substitution eliminates the mechanical movement entirely, thereby reducing power consumption, eliminating noise, and enabling faster processing.
Solution Approach 2:
The patent creates a virtual copy of the optical focusing process through digital computation. By capturing one image and generating multiple refocused images at different focal planes through image processing algorithms, the system replicates the effect of multiple physical captures without the actual mechanical movement, thus achieving depth information generation without power-intensive lens actuation.
2Measurement precision
If multiple cameras or image matrices are used to obtain depth information, then depth measurement capability is improved, but the image output size becomes small and processing complexity increases
Solution Approach 1:
The patent merges multiple images captured at different focal planes into a single composite image with extended depth of field. By combining the information from multiple images through digital processing, the system achieves both large output image size and accurate depth information without requiring complex multi-camera systems. This merging approach simplifies the overall system architecture while maintaining measurement precision.
3Measurement precision
If multiple small-sized images from multiple cameras are used, then depth information becomes easier to obtain, but the image resolution is low and generating clear large-sized 2D images becomes challenging
Solution Approach 1:
The patent transitions from the spatial dimension to the computational dimension by processing images digitally. Instead of relying on multiple physical cameras to provide both depth information and high resolution simultaneously, the system captures one high-resolution image and extracts depth information through computational analysis, then generates refocused images in the digital domain. This dimensional shift allows both high resolution and depth information to be achieved from a single capture.
Data Source
AI summary
An image processing device including an image sensor array, an image pre-processing unit, a depth information generator, and a focusing unit is provided. The image sensor array takes multiple images of a first object and a second object. The image pre-processing unit processes the images to generate two shift images associated with the two objects. The depth information generator generates depth information according to the two shift images. The depth information includes distance information associated with the first object. The focusing unit generates a pair of focused images that have the first object focused thereon according to the depth information and the two shift images.