Vehicular Image Colorization via Neural Network Reference Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Motor vehicles equipped with non-trichromatic image sensors face challenges in providing both high-resolution images for computer vision and color-accurate images for human viewing, leading to the need for multiple cameras and increased complexity and cost.
Innovation Solution
The system combines information from color and non-trichromatic sensors using image colorization algorithms and Convolutional Neural Networks to generate colorized images, utilizing color reference points from trichromatic cameras or databases to enhance non-RGB images from non-trichromatic sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If non-trichromatic image sensors are used, then resolution is improved, but color accuracy deteriorates
Solution Approach 1:
The patent merges data from non-trichromatic sensors (providing high-resolution luminance information) with color reference data from trichromatic cameras or databases to produce colorized images that maintain both high resolution and color accuracy for human viewing
Solution Approach 2:
The patent introduces color reference data as an intermediary element that bridges the gap between non-trichromatic sensor output and human color perception, using colorization algorithms to transfer color information from references to the high-resolution grayscale or non-trichromatic image
2Adaptability or versatility
If multiple cameras are deployed, then both computer vision and human viewing requirements are met, but device complexity increases
Solution Approach 1:
The patent makes non-trichromatic sensors serve dual functions by processing their high-resolution output for both computer vision applications and human viewing through colorization, eliminating the need for separate camera systems for each application
Solution Approach 2:
The patent combines data from multiple sources (non-trichromatic sensors, color references, databases) into a single unified output that satisfies both computer vision and human viewing requirements, reducing the total number of cameras needed
Data Source
AI summary
Apparatus for a motor vehicle comprises an image sensor generating non-trichromatic image data as an image pixel array covering a predetermined field of view relative to the vehicle. A color-calibration source generates at least one color reference value according to an object depicted in the non-trichromatic image data. A controller is configured to 1) associate the at least one color reference value to a set of pixels within the image pixel array, and 2) colorize the non-trichromatic image data according to the at least one color reference value to produce a colorized image. A display is configured to display the colorized image to a viewer in the vehicle.