Multi-Field Vision LIDAR Object Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional object recognition models struggle to accurately identify and track objects at varying distances due to limitations in image resolution, particularly when using wide view cameras that capture low-resolution images of objects beyond a certain threshold distance, while narrow view cameras improve resolution but reduce the field of view, limiting the number of objects that can be captured.
Innovation Solution
A vision-based LIDAR system employing multiple cameras with overlapping fields of view, combining wide view and narrow view images to enhance object recognition accuracy, using a distance sensor with a tracking beam to adjust the beam scanner and update object identification based on pixel locations and motion trajectories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If a wide view camera is used to capture images, then the field of view is widened to cover a large area of the environment, but the image resolution becomes low which limits the ability to distinguish and track objects beyond a threshold distance
Solution Approach 1:
The system divides the imaging task into multiple segments by using multiple cameras with different fields of view. A wide view camera captures the overall environment, while narrow view cameras capture high-resolution images of specific regions. This segmentation allows the system to simultaneously achieve wide coverage and high resolution without requiring a single camera to do both.
Solution Approach 2:
The system merges images from multiple cameras with different fields of view to create a composite view. The wide view image provides context and location information, while narrow view images provide high-resolution details. By combining these images and their corresponding depth maps, the system achieves both wide field of view and high measurement precision.
2Measurement precision
If a narrow view camera is used to improve image resolution, then the ability to distinguish objects at distance is improved, but the field of view becomes narrower which limits the number of objects that can be captured
Solution Approach 1:
The imaging system is segmented into multiple cameras, each responsible for a specific field of view. The wide view camera handles the broad area coverage, while narrow view cameras focus on specific regions requiring high resolution. This segmentation allows each camera to optimize for its specific function without compromising the overall system performance.
Solution Approach 2:
The system achieves multi-functionality by combining cameras with different fields of view. The wide view camera provides contextual information and object location, while narrow view cameras provide detailed inspection capability. Together, they create a universal imaging system that can both survey large areas and examine specific objects in detail.
3Measurement precision
If multiple cameras with different fields of view are used to simultaneously capture wide view and narrow view images, then object identification accuracy is improved, but the device complexity increases
Solution Approach 1:
The system merges multiple image sources and their corresponding depth maps into a unified representation. By combining the wide view image with narrow view images and their associated depth information, the system achieves high object identification accuracy while managing complexity through integrated processing.
Solution Approach 2:
The system uses an intermediary processing layer that receives images from multiple cameras, aligns them based on spatial relationships, and integrates their information. This intermediary layer manages the complexity of coordinating multiple cameras by providing a unified interface for image fusion and object identification.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system achieves improved object identification and tracking accuracy by integrating high-resolution details from narrow view images with contextual wide view images, enabling precise distance measurement and motion tracking of objects across different fields of view.
Implementation Method 1
a photo detector that detects at least a portion of the tracking beam that is reflected by the targeted object to measure the distance to the targeted object
Implementation Method 2
a vision based light detection and ranging (LIDAR) system
Data Source
AI summary
A vision based light detection and ranging (LIDAR) system captures images including a targeted object and identifies the targeted object using an object recognition model. To identify the targeted object, the vision based LIDAR system determines a type of object and pixel locations or a boundary box associated with the targeted object. Based on the identification, the vision based LIDAR system directs a tracking beam onto one or more spots on the targeted object and detects distances to the one or more spots. The vision based LIDAR system updates the identification of the targeted object based on the one or more determined distances.