Vehicle Camera Object Tracking via Movement Model Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing camera systems for motor vehicles face challenges in reliably capturing and tracking moved objects, such as pedestrians, due to their changing positions in consecutive images, especially when the vehicle or objects are in motion.

Innovation Solution

A method that recognizes a first object feature in a sequence of images, estimates its position using a movement model, predicts a feature in the next image, and associates it with a second object feature if a predetermined criterion is met, confirming the object's presence and tracking its movement.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Difficulty of detecting and measuring

If optical flow vectors are used to track objects in consecutive images, then movement detection capability is improved, but reliability of capturing moved objects deteriorates when objects change position significantly

Engineering Contradiction:

Improvemovement detection capabilityVSAvoidreliability of capturing moved objects

Core Design Contradiction:

Difficulty of detecting and measuringVSReliability

Solution Approach 1:

The system performs preliminary actions by predicting the position of moved objects in the second image based on the first object feature and estimated position before actually capturing the second image. This prediction creates a reference framework that guides the subsequent association process, allowing the system to proactively prepare for where objects should appear rather than passively reacting to their actual positions.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system employs feedback mechanisms by using the association result between the second object feature and prediction feature to confirm whether the object is truly moved. The association criterion acts as a feedback loop that validates the movement detection, comparing the predicted position with the actual detected position to reliably confirm object movement while filtering out false positives.

Inventive Principle:

Principle #23Feedback

2Duration of action of moving object

If object features are tracked across multiple images, then tracking capability is improved, but difficulty of differentiating objects from environment deteriorates

Engineering Contradiction:

Improvetracking capabilityVSAvoiddifficulty of differentiating objects from environment

Core Design Contradiction:

Duration of action of moving objectVSDifficulty of detecting and measuring

Solution Approach 1:

The system applies segmentation by dividing the image into multiple image cells and identifying characteristic features within each cell. This segmentation approach separates potential object features from the background environment, making it easier to track objects across images by focusing on discrete, identifiable features rather than trying to analyze the entire image at once.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The prediction feature acts as an intermediary between the first object feature and the second object feature. Instead of directly comparing features across images, the system uses the prediction feature (generated from movement models) as a mediator to bridge the gap, making the association process more reliable by providing an expected reference point that accounts for anticipated object movement.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Measurement precision

If movement models are used to estimate object positions, then prediction accuracy is improved, but device complexity deteriorates

Engineering Contradiction:

Improveprediction accuracyVSAvoiddevice complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system utilizes parameter changes by applying predetermined movement models that describe how object positions change over time. These models use parameters such as velocity and acceleration to estimate future positions, transforming the static object feature detection problem into a dynamic prediction problem that accounts for motion, thereby improving prediction accuracy through mathematical parameter transformations.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11170232B2Method for capturing an object in an environmental region of a motor vehicle with prediction of the movement of the object, camera system as well as motor vehicle

Publication Date: 2021.11.09 CONNAUGHT ELECTRONICS

AI summary

The invention relates to a method for capturing an object (9) in an environmental region (8) of a motor vehicle (1) based on a sequence of images (10, 11) of the environmental region (8), which are provided by means of a camera (4) of the motor vehicle (1), including the steps of: recognizing a first object feature (24) in a first image (10) of the sequence, wherein the first object feature (24) describes at least a part of the object (9) in the environmental region (8), estimating a position of the object (9) in the environmental region (8) based on a predetermined movement model, which describes a movement of the object (9) in the environmental region (8), determining a prediction feature (26) in a second image (11) following the first image (10) in the sequence based on the first object feature (24) and based on the estimated position, determining a second object feature (25) in the second image (11), associating the second object feature (25) with the prediction feature (26) in the second image (11) if a predetermined association criterion is satisfied, and confirming the second object feature (25) as originating from the object (9) if the second object feature (25) is associated with the prediction feature (26).