Fisheye Camera Object Detection via Reverse Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fisheye cameras offer a wider field of view but suffer from severe image deformation near the edges, making it difficult to accurately process and analyze images, particularly in intelligent traffic systems where lane lines and vehicle velocities are distorted, leading to discarded edge information.
Innovation Solution
The method involves projecting original images from a fisheye camera onto a cylindrical or spherical projection model, performing reverse mapping to create at least two images with different angles of view, allowing for object detection in both without cutting off edge information, thus fully utilizing the wide field of view.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If fisheye camera is used to capture images, then field of view is widened, but image deformation occurs at edges
Solution Approach 1:
The patent divides the fisheye image into multiple regions (central region and edge regions) and processes them differently. The central region is used for normal object detection, while edge regions are processed through reverse mapping to correct deformation, allowing the entire field of view to be utilized effectively.
Solution Approach 2:
The patent applies reverse mapping to transform the deformed edge regions of the fisheye image back to their original undistorted positions. This inversion process corrects the radial distortion by mapping pixels from the deformed image space back to the undistorted sensor space, enabling accurate object detection in previously unusable edge areas.
2Loss of information
If edge portion of fisheye image is retained, then wide field of view information is preserved, but object detection accuracy decreases due to deformation
Solution Approach 1:
The patent segments the image processing pipeline into different handling methods for central and edge regions. Central region images are processed directly, while edge region images undergo reverse mapping correction before object detection, thus preserving edge information while maintaining detection accuracy.
Solution Approach 2:
The patent changes the coordinate system and mapping parameters for edge region processing. By applying reverse mapping transformations that adjust pixel coordinates and spatial relationships, the system corrects deformation effects while retaining the expanded field of view information captured by the fisheye lens.
3Ease of operation
If conventional cameras are used for traffic surveillance, then image processing is easier, but field of view coverage is limited
Solution Approach 1:
The patent introduces reverse mapping as an intermediary processing step that bridges fisheye camera capabilities with conventional object detection algorithms. This intermediate transformation layer corrects fisheye distortion effects, allowing standard detection algorithms to work effectively on fisheye-captured images without requiring complete redesign of the processing pipeline.
Data Source
AI summary
This disclosure provides an apparatus and method for performing object detection based on images captured by a fisheye camera and an electronic device. The apparatus includes a memory and a processor coupled to the memory. The processor according to an embodiment is configured to: project an original image captured by the fisheye camera onto a cylindrical or spherical projection model, and perform reverse mapping to obtain at least two reversely mapped images, angles of view of the at least two reversely mapped images being towards different directions, detect objects in the reversely mapped images, respectively, and detect an object that is the same among the objects detected in the reversely mapped images. According to this disclosure, information in the wide field of view images obtained by capturing by the fisheye camera may be fully utilized.