Vehicle Camera Sensor Blockage Classification for Targeted Cleaning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting blockages in camera sensors used in vehicles are not precise enough, leading to unnecessary cleaning and potential delays in advanced driver assistance systems, as they rely on segmentation and comparison with reference images which can be challenging in real-world driving environments.
Innovation Solution
A method using a neural network algorithm to process camera images, determining the degree and class of blockage by segmentation and classification, and transmitting cleaning information to a cleaning device based on predetermined criteria, allowing for targeted and efficient cleaning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If segmentation and comparison with reference images are used to detect blockages, then blockage detection can be performed, but false detections occur and precision is insufficient
Solution Approach 1:
The patent replaces traditional image segmentation and comparison algorithms with a neural network-based machine learning system. The neural network is trained to recognize blockage patterns directly from camera images, substituting mechanical rule-based processing with adaptive intelligent processing that reduces false detections and improves precision.
Solution Approach 2:
The patent changes the detection parameters by using multiple blockage class categories (e.g., rain, snow, fog, dirt) with specific characteristics rather than a single binary blockage detection. The neural network analyzes multiple parameters simultaneously to determine both the presence and type of blockage, improving overall detection precision.
2Reliability
If cleaning is initiated for every detected blockage, then the camera sensor is kept clean, but unnecessary cleaning occurs wasting energy and cleaning media
Solution Approach 1:
The patent applies different cleaning actions based on the local quality of the detected blockage. Instead of uniform cleaning for all blockages, the system determines the specific blockage class (rain, snow, fog, dirt) and applies appropriate cleaning only when necessary, optimizing energy and cleaning media usage while maintaining sensor cleanliness.
Solution Approach 2:
The system uses feedback from the neural network's blockage classification to control the cleaning device. The cleaning decision is based on the analyzed blockage characteristics and degree, creating a feedback loop that prevents unnecessary cleaning actions and reduces energy consumption while maintaining optimal sensor performance.
3Measurement precision
If traditional segmentation methods are used, then blockage detection is possible, but processing time is increased and efficiency is reduced
Solution Approach 1:
The patent replaces computationally intensive traditional segmentation algorithms with a trained neural network that processes images more efficiently. The neural network has learned to identify blockage patterns during training, enabling faster real-time processing while maintaining or improving detection precision and efficiency.
4Adaptability or versatility
If camera sensor is placed at outer surroundings for ADAS functionality, then driver assistance capabilities are enabled, but the sensor becomes susceptible to blockage by weather and environmental factors
Solution Approach 1:
The system implements self-service by using the camera's own images to detect and diagnose blockages on its own lens or cover. The neural network analyzes the captured images to identify blockage patterns, enabling the system to self-diagnose and trigger appropriate cleaning actions without external intervention, maintaining ADAS functionality despite environmental exposure.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
Method for determining a cleaning information for an at least partially blocked camera sensor (3), which comprises a blockage on a transparent camera sensor component (6) in an optical path of the camera sensor (3), with the steps of: - Controlling the camera sensor (3) to capture at least one camera image, - Processing, by a computing device, the camera image (3) with a neural network algorithm, wherein the neural network algorithm is adapted to determine as an output a degree of camera sensor blockage by segmentation of at least a part of the camera image and a blockage class of a camera sensor blockage from a plurality of blockage classes by classification of at least a part of the camera image, - Determining a cleaning information in dependency of the degree of camera sensor blockage and the class of the camera sensor blockage, wherein the cleaning information describes that a cleaning of the camera sensor (3) is required, if a cleaning criterion is assigned to the determined class of the camera sensor blockage and if at least one degree threshold is exceeded by the determined degree of camera sensor blockage, and - Transmitting the cleaning information to a cleaning device (5) associated with the camera sensor (3) in order to clean the camera sensor (3) according to the cleaning information.