Pedestrian Intention Recognition Using Selective Image Sequence Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video-based systems for pedestrian detection and tracking are insufficient in recognizing a pedestrian's intention to cross the road, as existing detection and tracking algorithms fail to reliably predict the pedestrian's movement intention from their pose.
Innovation Solution
A method involving the evaluation of movement profiles from selectively chosen camera images using object detection and classification techniques, where a pedestrian is detected, and their movement profile is classified based on a sequence of images to determine their intended action, such as walking, standing, setting off, or stopping, using a classifier and pedestrian model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If detection and tracking algorithms are used to recognize pedestrians, then the position and speed of pedestrians can be established, but the algorithms are insufficient to recognize whether a pedestrian is intending to cross the road
Solution Approach 1:
The system performs preliminary classification of movement profiles by evaluating sequences of camera images to determine pedestrian intentions before actual crossing actions occur. The classifier analyzes movement patterns in advance to predict whether a pedestrian will cross, allowing proactive safety measures to be taken before the pedestrian actually enters the road.
2Reliability
If property vectors are formed and compared to reference vector clusters for pose recognition, then a pose representative of intended action can be associated with the road user, but the formation of property vectors is an elaborate process and actual intention cannot be established sufficiently reliably
Solution Approach 1:
The system extracts only the essential movement profile features from camera images that are relevant for intention recognition, rather than forming complex property vectors from all image data. The classifier selectively processes key movement characteristics to determine pedestrian intentions, simplifying the analysis process while maintaining reliability.
3Reliability
If all camera images in a sequence are evaluated for movement profile classification, then more comprehensive data is available, but the processing time and computational load increase
Solution Approach 1:
The system evaluates a selected subset of camera images from the sequence rather than processing all images. The classifier processes only the necessary number of images required to reliably determine movement profiles and pedestrian intentions, avoiding unnecessary computational overhead while maintaining classification accuracy.
Data Source
AI summary
A method and a driver assistance system recognize the intention of a pedestrian to move on the basis of a sequence of images of a camera. The method includes detecting a pedestrian in at least one camera image. The method also includes selecting a camera image that is current at the time t and selecting a predefined selection pattern of previous camera images of the image sequence. The method further includes extracting the image region in which the pedestrian was detected in the selected camera images of the image sequence. The method also includes classifying the movement profile of the detected pedestrian on the basis of the plurality of extracted image regions. The method outputs the class that describes the movement intention determined from the camera images of the image sequence.