Scene crop via adaptive view-depth discontinuity

Abstract

A method, apparatus, and system provide the ability to crop a three-dimensional (3D) scene. The 3D scene is acquired and includes multiple 3D images (with each image from a view angle of an image capture device) and a depth map for each image. The depth values in each depth map are sorted. Multiple initial cutoff depths are determined for the scene based on the view angles of the images (in the scene). A cutoff relaxation depth is determined based on a jump between depth values. A confidence map is generated for each depth map and indicates whether each depth value is above or below the cutoff relaxation depth. The confidence maps are aggregated into an aggregated model. A bounding volume is generated out of the aggregated model. Points are cropped from the scene based on the bounding volume.

Description

BACKGROUND OF THE INVENTION1. Field of the Invention[0001]The present invention relates generally to image processing, and in particular, to a method, system, apparatus, and article of manufacture for an adaptively cropping depth-based images.2. Description of the Related Art[0002]Images often contain distant or non-subject objects. The problem becomes how to effectively and efficiently crop such objects out of the images To better understand this problem, a description of image reconstruction and cropping may be useful.[0003]Three-dimensional scenes may often be constructed / reconstructed using various photogrammetry / point cloud processing techniques In particular, multiple cameras / point cloud scanners (e.g., laser scanners) may capture images of a scene that are then combined to create the 3D scene. Each captured image has associated depth data identifying the depth from the camera / scanner origin to the object in the scene There are often three different views that are captured: (1...

AI Technical Summary

Benefits of technology

The invention uses the depth map of each view of an image or camera to determine which parts of the image are a low confidence point. By combining all the views, the invention is able to accurately keep high confidence subject elements intact while removing distant objects. The technical effect of the invention is to improve the accuracy and reliability of image analysis and processing.

Problems solved by technology

However, while such a camera position-based cropping may work for Nadir shots, it is unable to correctly handle oblique or façade shots.

Specifically, because such a mechanism only relies on creating a convex hull from the XY projection of the camera position, it is unable to account for view-depth information resulting in over-cropping (only points near the camera are kept but not those the camera is actually looking at).

Images