Vehicle Collision Prediction Using Distance Change Trends
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current object collision prediction methods in intelligent driver assistance systems require significant computing power and storage space to identify the type of a detected target, leading to inefficiencies and errors, especially when data for the target is not stored in the database.
Innovation Solution
A method that predicts object collision by measuring distance changes between a vehicle and a detected target in images taken at different moments, eliminating the need to identify the target type, thereby reducing computational resources and storage requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the type of the to-be-detected target is identified based on image information and a prestored database, then the distance between the vehicle and the target can be estimated, but a large amount of computing power is consumed and significant storage space is required
Solution Approach 1:
The patent extracts only the essential feature needed for distance estimation - the size of the target in the image - while discarding the complex target type identification process. By taking out only the necessary information (target size ratio) and eliminating the database lookup step, the system achieves distance estimation without consuming large amounts of computing power or requiring significant storage space.
2Measurement precision
If the type of the to-be-detected target is identified based on image information and a prestored database, then the distance between the vehicle and the target can be estimated, but significant storage space is required to store the database
Solution Approach 1:
The patent removes the need for a prestored database by extracting only the target size information directly from the image. Instead of storing extensive target type data in a database, the system uses the ratio of the target's size in the image to a reference size, eliminating the need for large storage space while maintaining distance estimation capability.
3Measurement precision
If the type of the to-be-detected target is identified before distance estimation, then the distance can be calculated using real size data, but the process consumes a large amount of computing power
Solution Approach 1:
The patent extracts only the necessary size information from the image and directly calculates distance using the size ratio method, removing the time-consuming target type identification step. This extraction approach maintains distance estimation accuracy while significantly improving collision prediction efficiency by eliminating unnecessary processing steps.
Data Source
AI summary
This application provides a collision detection method and related apparatus. An image taken by a photographing unit may be used to predict whether a collision with a to-be-detected target will occur. In a current collision prediction method, a type of the to-be-detected target needs to be determined first based on the image taken by the photographing unit, which requires consuming of a large amount of computing power. In the collision prediction method provided in this application, a change trend of a distance between the to-be-detected target and a vehicle in which the apparatus is located may be determined based on the distances between the to-be-detected target and the vehicle at different moments, to predict a collision between the to-be-detected target and the vehicle. This method can improve efficiency in collision prediction and reduce energy consumption in predicting collision.