Object Detection Apparatus Initial Direction Vector Convergence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection systems face challenges in converging predicted values for the traveling direction of objects, requiring multiple detection results and delaying calculations when initial direction vectors are not set correctly.
Innovation Solution
An object detection apparatus that acquires and processes object information to set multiple initial direction vectors for predicting object positions, allowing for early convergence of traveling direction predictions by associating with the closest actual direction vector.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple initial direction vectors are set for predicting object positions, then the convergence of predicted values for traveling direction is improved, but the device complexity increases
Solution Approach 1:
The prediction process is segmented into multiple parallel prediction paths, each using a different initial direction vector. This allows the system to explore multiple possible traveling directions simultaneously, improving convergence by selecting the best matching prediction from multiple candidates rather than relying on a single prediction path.
Solution Approach 2:
Multiple initial direction vectors are set in advance before actual object tracking begins. This preliminary setup of multiple potential direction vectors allows the prediction unit to immediately start multiple prediction processes without waiting for convergence data, thereby improving the convergence speed of predicted values.
2Loss of time
If multiple initial direction vectors are set for predicting object positions, then the calculation time is reduced through early convergence, but the processing complexity increases
Solution Approach 1:
Multiple initial direction vectors are pre-configured in the initial setting unit based on the detection result. This preliminary action eliminates the need for iterative calculation to determine the first prediction direction, allowing the system to immediately proceed with multiple parallel predictions and achieve early convergence, thereby reducing overall calculation time.
Solution Approach 2:
The system dynamically selects the most appropriate initial direction vector from multiple candidates based on the actual object behavior. By maintaining multiple dynamic prediction paths and selecting the one that converges fastest or matches actual object movement best, the system reduces calculation time while managing processing complexity through adaptive selection.
3Measurement precision
If regression analysis is used to calculate traveling direction and speed, then the accuracy of object tracking is improved, but the requirement for multiple detection results within predetermined time increases processing delays
Solution Approach 1:
Instead of waiting to collect multiple detection results before performing regression analysis, the system performs preliminary action by setting multiple initial direction vectors based on the first detection result. This allows prediction to begin immediately with multiple potential directions, eliminating the waiting time required to accumulate sufficient data for regression analysis while maintaining tracking accuracy through subsequent convergence validation.
Data Source
AI summary
A vehicle information providing apparatus repeatedly acquires observation point information. The apparatus calculates at least one position of the predicted point. When the observation point information is acquired, the apparatus extracts a predicted point that can be connected to the observation point from the calculated predicted points. The apparatus calculates the position of the current filtered point based on the position of the current observation point and the position of the latest predicted point. When the observation point information of a new object (an object having no latest predicted point connectable to the observation point) is acquired, the device sets initial vectors. The apparatus calculates the position of the predicted point at time instants succeeding the current time instant for each of the traveling directions indicated by the initial vectors.