Fish-Eye Image Normalization for Vehicle Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection systems using fish-eye cameras struggle with accurate detection of approaching objects due to distortion, particularly in determining actual distance and speed, which is crucial for safety assistance in vehicle parking scenarios.
Innovation Solution
An object detection apparatus and method that normalizes fish-eye images by acquiring and processing viewpoint compensation vectors to correct distortion, allowing for accurate determination of object position and movement relative to the vehicle.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If a fish-eye camera is used to obtain a wide field of view at low cost, then the field of view coverage is improved, but the detection accuracy of object distance and speed deteriorates due to distortion
Solution Approach 1:
The patent transforms the fish-eye image parameters by applying distortion correction algorithms that map the distorted coordinates to corrected coordinates based on the fish-eye lens characteristics. This parameter transformation enables accurate distance and speed measurement while maintaining the wide field of view advantage of the fish-eye camera.
Solution Approach 2:
The patent introduces an intermediary processing step that includes detecting the fish-eye lens parameters, calculating distortion correction coefficients, and applying these corrections to the image data. This intermediary processing layer bridges the gap between the distorted fish-eye image and the accurate measurement requirement.
2Device complexity
If object detection is performed based on the size of the target object in the fish-eye image, then the detection process is simplified, but the accuracy of determining actual distance and moving speed deteriorates
Solution Approach 1:
The patent changes the detection parameters by using distortion-corrected image coordinates instead of raw fish-eye image coordinates. The correction process transforms the size-based detection approach into an accurate distance and speed measurement approach by applying the fish-eye lens distortion model to relate image coordinates to real-world measurements.
3Ease of operation
If the determination of approaching moving body is based on the size increase in the fish-eye image, then the detection method is simple, but the reliability of risk estimation deteriorates especially for peripheral objects
Solution Approach 1:
The patent transforms the simplicity-based detection method into a reliability-based method by changing the parameter from image size alone to distortion-corrected position and velocity vectors. The correction process accounts for the fish-eye lens distortion characteristics, enabling reliable risk estimation for objects at all positions in the field of view, including peripheral regions.
Data Source
AI summary
An object detection apparatus 100 is an apparatus for detecting an object in a fish-eye image. The object detection apparatus 100 includes a normalized image acquisition unit 10 configured to acquire a normalized image obtained by normalizing a fish-eye image in which an object appears; a position detecting unit 20 configured to detect position coordinates of the object in the normalized image; and a determination unit 30 configured to determine a positional relationship between the object and the object detection apparatus 100 using the position coordinates of the object in the normalized image.