Tilted Sensor Depth Map Generation for Artwork

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current high-resolution imagery fails to capture depth information, particularly in artwork, due to the loss of texture details and concerns about laser radiation from LIDAR technology.

Innovation Solution

A method and system that utilize a tilted sensor plane to capture images with an extremely shallow depth of field, allowing for software focus detection and generation of depth maps by analyzing in-focus regions and assigning position data to create a comprehensive depth map, while avoiding the use of LIDAR.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If LIDAR is used to create depth maps, then depth information can be captured, but laser radiation may harm artwork

Engineering Contradiction:
Improvedepth informationVSAvoidlaser radiation
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent replaces the optical/LIDAR system with a mechanical camera system that uses focused light capture. Instead of projecting laser beams to measure depth, the system uses a camera with a tilted sensor plane to capture images at different focus depths, substituting active optical measurement with passive optical capture and computational analysis.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces an intermediary computational process that analyzes image focus characteristics to infer depth information. Rather than directly measuring depth with laser, the system uses software to detect which regions are in focus in each image and translates this focus information into depth map data, serving as an indirect measurement intermediary.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If conventional photography is used to capture high resolution images, then image resolution is improved, but depth information and texture details are lost

Engineering Contradiction:
Improveimage resolutionVSAvoiddepth information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent segments the imaging process into multiple focused captures. By dividing the scene into different depth planes and capturing images with selective focus at each plane, the system obtains multiple images where different regions are sharp. This segmentation of focus across multiple captures allows reconstruction of depth information that would be lost in a single conventional photograph.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent adds a depth dimension to conventional 2D photography by introducing the tilt angle between sensor plane and object plane. This geometric configuration creates a relationship between focus position and depth, transforming standard 2D image capture into a 3D-aware imaging system that preserves depth information alongside high-resolution texture details.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This approach effectively captures and represents depth information of objects, enhancing the detail and accuracy of high-resolution images without the risks associated with LIDAR, and can be used in multi-camera systems to create global depth maps efficiently.

Implementation Method 1

capture a plurality of source images using a sensor plane that is tilted with respect to an object plane and an associated depth of field plane

Methodology Applied
Scientific EffectDepth of Field: Depth of Field

Data Source

PatentEP2826243B1Method and system for identifying depth data associated with an object
Publication Date: 2020.06.10 GOOGLE LLC
  • EP2826243B1 patent drawingFigure 1
  • EP2826243B1 patent drawingFigure 1A
  • EP2826243B1 patent drawingFigure 2

AI summary

Methods and systems for identifying depth data associated with an object are disclosed. The method includes capturing, with an image capturing device, a plurality of source images of the object. The image capturing device has a sensor that is tilted at a known angle with respect to an object plane of the object such that the image capturing device has a depth of field associated with each source image, the depth of field defining a plane that is angled with respect to the object plane. An image processor analyzes the plurality of source images to identify segments of the source images that satisfy an image quality metric. Position data is assigned to the identified segments of the source images, the position data including depth positions based on the plane defined by the depth of field.