Convolutional Network Image Enhancement via Multi-Resolution Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image segmentation in computer graphics is challenging due to sensitivity to image scale, lighting, and location changes, particularly in deep learning model-based methods.
Innovation Solution
An image enhancement device with a down-sampling module, correction modules, and an up-sampling module is integrated into a convolutional network apparatus, which down-samples, corrects, and up-samples images using correction models and parameters to generate improved images, enhancing segmentation results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If deep learning model-based segmentation method is used, then segmentation can be performed automatically, but segmentation result is affected by image scale, lighting and location changes
Solution Approach 1:
The patent applies preliminary action by performing image enhancement operations (down-sampling, correction, up-sampling) on the input image before it enters the deep learning segmentation network. The correction modules adjust illumination, contrast, and other parameters in advance to ensure the image meets optimal input conditions for the segmentation model, thereby improving segmentation reliability without compromising automation.
2Measurement precision
If image correction is performed on down-sampled images, then segmentation accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent segments the image processing task into distinct modules: down-sampling module, multiple correction modules (illumination correction, contrast correction, etc.), and up-sampling module. Each module performs a specific function, making the complex processing pipeline more manageable and maintainable while achieving improved segmentation accuracy through coordinated operation of these specialized components.
Solution Approach 2:
The correction modules are designed with multi-functionality to handle various image quality issues (illumination variations, contrast problems, noise) within a unified framework. This universal approach reduces overall system complexity compared to having separate specialized processors for each type of image defect.
Data Source
AI summary
An image enhancement device that includes a down-sampling module, correction modules and an up-sampling module is provided. The down-sampling module down-samples an input image to generate down-sampled images having different down-sampled resolutions. Each of the correction modules performs correction on one of the down-sampled images according to a correction model based on at least one correction parameter to generate one of corrected images. The up-sampling module up-samples the corrected images to generate up-sampled images, wherein each of the up-sampled images is of a same up-sampled resolution. The concatenating module concatenates the up-sampled images into an output image.