Camera Object Detection via Sensor-Guided Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection systems in vehicles face challenges in efficiently processing information from cameras and other sensors due to high computational demands, which can exceed the processing capabilities of economically viable on-board processors.
Innovation Solution
An object detection system that utilizes information from a camera and other detectors like LIDAR or radar to select and prioritize portions of the camera output based on object presence, dividing these into segments and patches to determine Objectness, thereby reducing computational load while maintaining accuracy in object detection and tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the processor analyzes the entire camera output to detect objects, then detection accuracy is improved, but processing time and computational load increase significantly
Solution Approach 1:
The camera output is divided into multiple segments based on LIDAR detection data. Instead of processing the entire camera image, the system identifies and processes only those segments that contain potential objects detected by the LIDAR sensor. This segmentation approach maintains detection accuracy for relevant areas while significantly reducing the overall processing time and computational load.
Solution Approach 2:
The system extracts and processes only the relevant portions of the camera output that correspond to LIDAR-detected objects. By taking out and focusing computational resources on specific regions of interest rather than the entire image, the system achieves accurate object detection while minimizing processing time and resource consumption.
2Measurement precision
If the processor analyzes the entire camera output to detect objects, then detection accuracy is improved, but computational load exceeds processor capabilities
Solution Approach 1:
The camera output is divided into multiple segments based on LIDAR detection data. Instead of processing the entire camera image, the system identifies and processes only those segments that contain potential objects detected by the LIDAR sensor. This segmentation approach maintains detection accuracy for relevant areas while significantly reducing the overall processing time and computational load.
Solution Approach 2:
The system extracts and processes only the relevant portions of the camera output that correspond to LIDAR-detected objects. By taking out and focusing computational resources on specific regions of interest rather than the entire image, the system achieves accurate object detection while minimizing processing time and resource consumption.
3Productivity
If the system processes only selected portions of camera output, then processing efficiency is improved, but risk of missing objects increases
Solution Approach 1:
The system merges the detection capabilities of two different sensor types - LIDAR for initial object detection and camera for detailed analysis. By combining the strengths of both sensors, the system achieves high processing efficiency through selective camera processing while maintaining high reliability through the complementary nature of multi-sensor detection.
Solution Approach 2:
The LIDAR sensor acts as an intermediary that guides the camera processing. The LIDAR detects potential objects first, and its data serves as a mediator to select which portions of the camera output require detailed processing. This intermediary approach ensures that no objects are missed while maintaining processing efficiency.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
An illustrative example object detection system includes a camera (104) having a field of view. The camera (104) provides an output comprising information regarding potential objects within the field of view. A processor (106) is configured to select a portion of the camera output based on information from at least one other type of detector (110) that indicates a potential object in the selected portion. The processor determines an Objectness of the selected portion based on information in the camera output regarding the selected portion.