Panoramic Camera Target Elimination via Image Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Panoramic cameras, due to their wide angle of view, often capture undesired objects such as photographers or small animals, leading to errors during 3D modeling, necessitating a method to efficiently eliminate target objects while maintaining image quality.
Innovation Solution
A target object elimination method using a panoramic camera with at least two independent lenses, involving image capture, target recognition, division of images into parts, deletion of target-containing parts, and combination of remaining parts to produce a corrected image without the target object, employing both artificial and machine recognition techniques and pixel-level color averaging.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If a panoramic camera is used to capture wide-angle scenes, then the coverage area is improved, but the number of undesired objects captured increases
Solution Approach 1:
The patent divides the panoramic image into multiple independent image parts (e.g., left half and right half) based on the position of undesired objects. This segmentation allows selective processing of only the affected regions while preserving the rest of the image, thereby maintaining wide coverage while removing harmful elements.
Solution Approach 2:
The patent extracts and removes undesired objects from the panoramic image by identifying their locations and eliminating the corresponding image parts. This extraction process preserves the main scene coverage while removing harmful elements such as photographers, pedestrians, or animals that are not part of the intended 3D modeling subject.
2Manufacturing precision
If target objects are eliminated from panoramic images, then image quality is improved, but processing complexity increases
Solution Approach 1:
By dividing the panoramic image into discrete image parts, the patent simplifies the removal process. Instead of processing the entire large image at once, the system works with smaller, manageable segments, reducing computational complexity while maintaining processing thoroughness.
Solution Approach 2:
The patent applies partial action by processing only the specific image parts that contain undesired objects rather than the entire image. This selective processing reduces the overall computational load while ensuring that image quality is improved in the critical areas where errors would occur.
Data Source
AI summary
The present invention provides a target object elimination method based on a panoramic camera, the panoramic camera comprising at least two independent lenses, wherein the method comprises following steps: capturing one basic panoramic image by each of the independent lenses; recognizing a target object from each of the basic panoramic images; dividing each of the basic panoramic images into a plurality of image parts having a same or different size, wherein at least one of the image parts comprises the target object; deleting the one image part comprising the target object in each of the basic panoramic images; and combining all remaining image parts in each of the basic panoramic images to obtain a corrected panoramic image from which the target object is eliminated. By the present invention, the target object can be eliminated quickly and efficiently while ensuring the quality of the image.