Document Detection via Internal Feature Projection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies face challenges in capturing and processing documents using mobile devices due to limited processing power, image resolution, and the inability to accurately detect objects with missing or obscured edges in image and video data.

Innovation Solution

A computer-implemented method and system for detecting documents in digital images using mobile devices, which involves detecting internal features of the object and projecting the location of object edges based on these features, allowing for accurate detection and reconstruction of objects in a three-dimensional coordinate space.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional scanner-based processing algorithms are used on mobile devices, then document detection accuracy is improved, but processing time and computational cost become prohibitively expensive

Engineering Contradiction:
Improvedocument detection accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent extracts and focuses only on the most critical processing steps needed for document detection, removing unnecessary computational overhead from conventional scanner-based algorithms. This allows the system to achieve adequate detection accuracy with significantly reduced processing time suitable for mobile devices.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent modifies processing parameters and thresholds to optimize performance for mobile device constraints. By adjusting detection sensitivity, processing resolution, and algorithmic parameters, the system achieves acceptable document detection accuracy while maintaining fast processing speeds on mobile platforms.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If conventional scanner-based processing algorithms are used on mobile devices, then document detection accuracy is improved, but device complexity and processing power requirements increase

Engineering Contradiction:
Improvedocument detection accuracyVSAvoidprocessing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent employs simpler, more lightweight processing algorithms that can be executed on mobile devices with limited resources. Rather than implementing complex conventional algorithms, the system uses optimized, simplified versions that reduce computational complexity while maintaining adequate detection accuracy for mobile applications.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

Solution Approach 2:

The patent divides the document detection process into distinct segments or stages, each handling specific aspects of detection. This segmentation allows the system to process images in manageable steps, reducing overall computational complexity and making the algorithm more suitable for mobile device constraints.

Inventive Principle:
Principle #1Segmentation

3Ease of operation

If mobile capture components are used, then portability and ease of operation are improved, but image quality and consistency deteriorate due to projective effects and distortions

Engineering Contradiction:
ImproveportabilityVSAvoidimage quality consistency
Core Design Contradiction:
Ease of operationVSManufacturing precision

Solution Approach 1:

The patent converts the projective effects and distortions introduced by mobile camera capture into beneficial information. By detecting and analyzing these distortions, the system can calculate the three-dimensional orientation and position of the document, then apply corrective transformations to produce a rectified, high-quality image suitable for further processing.

Inventive Principle:
Principle #22Blessing in disguise (Convert harm into benefit)

Solution Approach 2:

The patent addresses two-dimensional image distortions by introducing three-dimensional geometric reconstruction. By modeling the document in 3D space and considering camera orientation, position, and focal length, the system can computationally rectify the captured image to correct perspective distortions and produce a consistent, high-quality output.

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

4Measurement precision

If edge-based detection methods are used, then object boundary detection is improved, but detection reliability deteriorates when edges are missing or obscured

Engineering Contradiction:
Improveedge detection accuracyVSAvoiddetection reliability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent uses internal features rather than requiring complete edge detection. By focusing on detecting sufficient internal features (such as text lines, shapes, or patterns) rather than attempting to detect all edges, the system achieves reliable document detection even when edges are partially missing or obscured in the captured image.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3436865B1Content-based detection and three dimensional geometric reconstruction of objects in image and video data
Publication Date: 2025.06.18 TUNGSTEN AUTOMATION CORPORATION
  • EP3436865B1 patent drawingFigure 1
  • EP3436865B1 patent drawingFigure 2
  • EP3436865B1 patent drawingFigure 3A~3B

AI summary

Systems, computer program products, and techniques for detecting objects depicted in digital image data are disclosed, according to various exemplary inventive concepts. The inventive concepts utilize internal features to accomplish object detection, thereby avoiding reliance on detecting object edges and/or transitions between the object and other portions of the digital image data, e.g. background textures or other objects. The inventive concepts provide an improvement over conventional object detection since objects may be detected even when edges are obscured or not depicted in the digital image data. In one aspect, a computer-implemented method of detecting an object depicted in a digital image includes: detecting a plurality of identifying features of the object, the plurality of identifying features being located internally with respect to the object; and projecting a location of one or more edges of the object based at least in part on the plurality of identifying features.