Driver Mobile Phone Detection Using Side-Window CNN Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting mobile device usage by drivers in traffic accidents are inefficient, relying heavily on human observation and requiring significant manpower, resulting in a low documentation and enforcement of such violations.
Innovation Solution
A machine learning algorithm, specifically a Convolutional Neural Network (CNN), is trained using captured images from a road environment to classify drivers using or not using a mobile device, optimizing the training process through iterative adjustments of image sets and parameters to improve accuracy and minimize false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If human operators manually review captured video to detect mobile device usage by drivers, then detection capability is achieved, but considerable manpower is required and only a small fraction of violations are documented
Solution Approach 1:
The patent replaces the mechanical system of human operators manually reviewing video with an automated machine learning system. A neural network is trained to automatically analyze captured images and video frames, identify drivers using mobile devices, and generate detection results without human intervention. This substitution eliminates the bottleneck of manual review while maintaining detection capability, thereby dramatically increasing the fraction of violations that can be documented and enforced.
2Reliability
If conventional manual monitoring methods are used, then mobile device usage violations can be detected, but significant manpower resources are consumed
Solution Approach 1:
The system enables self-service detection where the machine learning algorithm automatically performs the entire detection process without requiring human operators. The neural network independently analyzes video frames, identifies violations, and generates results. This self-service capability maintains reliable violation detection while eliminating the need for significant manpower resources, as the automated system handles all detection tasks autonomously.
3Measurement precision
If more training images are added to improve machine classification accuracy, then true positive detection rate increases, but training time and computational resources increase
Solution Approach 1:
The patent applies preliminary action by carefully curating and preparing a diverse set of training images before initiating the training process. The training dataset is pre-selected to include various scenarios, lighting conditions, angles, and types of mobile device usage. This preliminary preparation ensures that the neural network learns from comprehensive examples, achieving high classification accuracy. By optimizing the training dataset composition beforehand, the system maximizes detection performance while minimizing the time and computational resources required during actual training execution.
Data Source
AI summary
Determining that a motor vehicle driver is using a mobile device while driving a motor vehicle. Multiple images of a driver of a motor vehicle are captured through a side window of the motor vehicle. Positive images show a driver using a mobile device while driving a motor vehicle. Negative images show a driver not using a mobile device while driving a motor vehicle. Multiple training images are selected from both the positive images and the negative images. The selected training images and respective labels, indicating that the selected training images are positive images or negative images, are input to a machine (e.g. Convolutional Neural Network, (CNN)). The CNN is trained to classify that a test image, captured through a side window of a motor vehicle, shows a driver using a mobile device while driving the motor vehicle.