Compressed Image Capture Using DMD Wavelet Coefficients
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional image capture methods require large memory and computational resources to store and compress images, as they first capture all image pixels and then compress them, which is inefficient and complex.
Innovation Solution
The method captures images directly in a compressed domain by computing wavelet coefficients using a digital micro-mirror device (DMD) module, which reflects specific portions of the image onto a detector, reducing the need for full-image sampling and allowing adaptive quality control and low-complexity decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional image capture methods are used to store and compress images, then complete image data is captured, but memory and computational resources are significantly increased
Solution Approach 1:
The patent extracts only the essential image information by directly capturing wavelet coefficients in the compressed domain, rather than capturing complete pixel data. This is achieved through a DMD module that selectively reflects specific portions of the image onto a detector, extracting only the necessary measurements for image reconstruction while discarding redundant information.
Solution Approach 2:
The patent changes the measurement parameter from pixel intensity values to wavelet coefficients. By transforming the capture process to directly measure wavelet coefficients through the DMD module's selective reflection and the detector's measurement of weighted sums, the system captures image data in a compressed representation that requires fewer resources for storage and processing.
2Loss of information
If all image pixels are captured first, then complete image information is obtained, but capture time and processing complexity are increased
Solution Approach 1:
The patent performs preliminary compression during the capture process itself rather than after capture. The DMD module is pre-configured to reflect only the specific portions of the image that contribute to the wavelet coefficients being measured, so the compression action is built into the capture mechanism, eliminating the need for separate compression steps.
Solution Approach 2:
The patent replaces the traditional mechanical/optical capture system that records all pixels with a transformed system using a DMD module and photodetector array. This substitution enables direct measurement of wavelet coefficients through optical weighting and detection, fundamentally changing how image data is acquired from the ground up.
3Reliability
If traditional compression methods are applied after capture, then image quality is maintained, but device complexity and processing requirements are increased
Solution Approach 1:
The patent inverts the traditional workflow by capturing images directly in the compressed domain rather than capturing full-resolution images and then compressing them. The DMD module and detector system are configured to directly measure wavelet coefficients, which are the compressed representation, thereby inverting the conventional capture-compress sequence into compress-capture.
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 approach significantly reduces the time and complexity of capturing compressed images, enables adaptive quality adjustment, and facilitates efficient reconstruction of the original image from wavelet coefficients, using substantially fewer measurements than traditional methods.
Implementation Method 1
an image is captured or acquired by projecting the image onto one or more photodetectors, which convert light into a current or voltage
Implementation Method 2
a mirror array module, a mirror module controller to control the mirror array module to reflect a first plurality of portions of the image onto a detector
Data Source
AI summary
Example methods and apparatus to capture compressed images are disclosed. A disclosed example method includes capturing a first output of a first photodetector representative of a first weighted sum of the a plurality of portions of an image, capturing a second output of a second photodetector representative of a second weighted sum of a second plurality of portions of the image, and computing a first wavelet coefficient for the image using the first and second captured outputs.