Onboard Camera Image Processing for Standing Object Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In onboard camera systems, standing objects appear collapsed in viewpoint transformation images due to affine transformation assumptions, making it difficult for users to recognize them as solid objects.
Innovation Solution
An image processing device and method that performs movement transformation to adjust subject positions based on their distance from a vanishing point, ensuring standing objects are displayed in an upright manner.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If affine transformation is used for viewpoint transformation, then the transformation can be performed under the assumption that subjects exist in a plane or curve, but standing objects are transformed into collapsed images that are difficult to recognize
Solution Approach 1:
The image processing is divided into two distinct stages: first, affine transformation is applied to the entire image for efficient viewpoint transformation; second, detected standing objects are extracted and re-transformed using perspective transformation. This segmentation allows each transformation method to be applied appropriately, maintaining both computational efficiency and object recognition quality.
Solution Approach 2:
Different transformation methods are applied to different regions of the image based on object type. Standing objects detected in the affine-transformed image are subjected to localized perspective transformation, while the rest of the image retains the affine transformation results. This local quality approach ensures that standing objects are correctly represented without compromising the overall transformation efficiency.
2Ease of operation
If perspective transformation is applied to correct standing object distortion, then object recognition is improved, but the transformation becomes more complex and computationally intensive
Solution Approach 1:
The affine transformation is performed first as a preliminary step to establish the viewpoint transformation for the entire image. This preliminary action creates a base transformed image that can then be selectively adjusted. By performing the simpler affine transformation first, the system avoids the computational burden of applying perspective transformation to the entire image while still achieving correct standing object representation.
Solution Approach 2:
The transformation approach is made dynamic by detecting standing objects in the affine-transformed image and applying perspective transformation only to those specific regions. This dynamic adaptation allows the system to adjust the transformation complexity based on the actual content of the image, applying the more computationally intensive perspective transformation only where necessary for standing objects.
Data Source
AI summary
The present technology relates to an image processing device, an image processing method, and a program capable of making it easier to recognize standing objects.Movement transformation of moving a subject position where a subject appears in a target image to be processed is performed, depending on a subject distance from a vanishing point in the target image to the subject position. The present technology can be applied to, for example, the image processing and the like of an image taken by a camera unit onboard a vehicle or other moving body.