Time-to-Angle Camera Sorting for Clear Imaging Through Fog
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Camera-based object recognition systems struggle to effectively distinguish echo signals from fog or rain droplets from object signals in adverse weather conditions, limiting their effectiveness in autonomous driving and aerial mobility applications.
Innovation Solution
A time-to-angle sorting system using a laser, digital micromirror device, and optical imagers to diffract and time-stamp return signals from fog and object reflections, enabling separation of these signals for clear imaging through fog or rain.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a camera-based system is used for object recognition, then resolution and signal contrast are improved in dense traffic, but performance deteriorates in bad weather conditions (fog, rain, snow)
Solution Approach 1:
The patent segments the return signal into multiple time intervals corresponding to different distances. By dividing the temporal profile into segments, the system can distinguish between near-range fog droplets and far-range objects based on their respective echo return times, enabling separation of harmful fog signals from useful object signals.
Solution Approach 2:
The patent transforms the one-dimensional temporal signal into a two-dimensional representation by mapping time intervals to spatial distances. This dimensional transformation allows the system to visualize and process fog versus object reflections in a structured manner, improving discrimination capability.
2Object-affected harmful factors
If time-of-flight lidar is applied to camera-based SA, then ability to see through fog is improved, but echo signals from water droplets obscure the signal of interest
Solution Approach 1:
The patent extracts and removes the harmful fog echo components from the total return signal by identifying them through their characteristic time profiles. By taking out the near-range fog signals, the system isolates the far-range object signals that would otherwise be obscured.
Solution Approach 2:
The patent converts the harmful effect of water droplet reflections into a beneficial discrimination mechanism. The echoes from fog droplets, while initially obstructive, provide temporal information about their distance, which can be used to identify and exclude them from the measurement, thereby improving object detection.
3Object-affected harmful factors
If radar is used for situation awareness, then ability to see through fog is improved due to longer wavelength, but resolution and signal contrast deteriorate for small objects
Solution Approach 1:
The patent merges the complementary strengths of camera and radar systems. It combines the camera's high resolution for small objects with the radar's ability to penetrate fog, using the temporal profile information from laser pulses to achieve both fog penetration and fine detail detection simultaneously.
Solution Approach 2:
The patent creates a multi-functional sensing system that can perform multiple functions: detecting small objects with high resolution, penetrating fog, and providing 3D spatial information. The single laser-based system achieves what would traditionally require multiple specialized sensors.
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
Enhances camera-based object recognition in adverse weather by separating fog and object signals, providing clear images and improving situational awareness in autonomous systems.
Implementation Method 1
a laser oriented to direct a series of pulses of light in a predetermined direction
Implementation Method 2
receive a return signal from an object, from any of a first series of water droplets that may be positioned between the laser and the object
Implementation Method 3
switching each micromirror from an on-state through a transitional period to an off-state so that the return signal is sequentially diffracted into a first diffraction order representing a first portion of the return signal reflected by the first series of water droplets
Implementation Method 4
an array of optical imagers positioned to capture a series of images over time of each of the first diffraction order, the second diffraction order, and the third diffraction order
Implementation Method 5
map a time of arrival of each the series of images to the first diffraction order, the second diffraction order, and the third diffraction order to separate the second portion of the return signal from the first portion of the return signal
Data Source
AI summary
A time to angle camera sorting system that uses a short pulsed laser, a lens, an array of DMDs, a camera, and a processor. As the laser illuminates the space before it, DMDs turn on and off between their tilt angles, and diffracted light is returned and captured. The light data is sent to the processors which are programmed, configured, or structured to map the time of arrival of the light data with the angle data of the DMDs associated with each light data point, and then process an image based on the mapped time to angle mapped data.