3D Object Recognition in Augmented Reality Using Characteristic Measurements
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current augmented reality engines struggle with recognizing three-dimensional objects and maintaining alignment when the camera moves relative to the object, especially for objects with complex shapes or without markers, leading to inconsistent and inaccurate overlays.
Innovation Solution
A method that uses characteristic measurements from a three-dimensional drawing to align the augmented reality content with the real object, allowing the electronic device to move freely while maintaining alignment using sensors like IMU and touch interface adjustments, eliminating the need for image processing and marker detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image recognition systems based on camera detection are used, then two-dimensional objects and regular three-dimensional objects can be recognised, but three-dimensional objects with complex shapes cannot be recognised and alignment is lost when the camera moves relative to the object
Solution Approach 1:
The patent transitions from two-dimensional image recognition to three-dimensional spatial recognition by using characteristic measurements (landmarks) defined in 3D space from technical drawings. These landmarks have spatial coordinates (x, y, z) that enable the system to recognize and track three-dimensional objects with complex shapes, not just flat images or regular objects.
Solution Approach 2:
The system performs preliminary action by pre-defining characteristic measurements and landmarks from three-dimensional technical drawings before the actual recognition process. These landmarks serve as reference points that are established in advance, enabling the augmented reality system to quickly and accurately locate and align with the target object without requiring complex real-time image processing.
2Reliability
If markers are used for augmented reality activation, then augmented reality can be activated, but the marker must be continuously viewed and the system remains marker-dependent
Solution Approach 1:
The patent extracts the essential recognition function from the marker itself and transfers it to the three-dimensional object's inherent geometric features (landmarks). By removing the marker dependency, the system uses the object's own characteristic measurements as the basis for recognition and alignment, eliminating the need for separate markers while maintaining reliable augmented reality activation.
Solution Approach 2:
The three-dimensional object serves itself for recognition purposes by providing its own characteristic measurements and landmarks as reference points. The object's geometry inherently contains the necessary information for identification and alignment, eliminating the need for external markers or labels.
3Adaptability or versatility
If extended markers are used to allow camera movement, then augmented reality content remains available, but alignment between computer-generated image and real object is lost
Solution Approach 1:
The system dynamically updates the position and orientation of the augmented reality content based on the camera's movement relative to the three-dimensional landmarks. By continuously tracking the spatial relationship between the camera and the fixed landmarks on the object, the system maintains accurate alignment even as the camera moves, unlike static marker-based systems.
Solution Approach 2:
The system uses feedback from the detected positions of characteristic measurements and landmarks to continuously adjust and correct the alignment of the augmented reality content. This closed-loop approach ensures that the computer-generated image remains properly aligned with the real object during camera movement by constantly comparing the expected landmark positions with the actual camera viewpoint.
4Measurement precision
If image processing and marker detection are used for object recognition, then recognition can be achieved, but the process becomes complex and time-consuming
Solution Approach 1:
The system performs preliminary action by pre-defining characteristic measurements and landmarks from three-dimensional technical drawings before the actual recognition process. These landmarks serve as reference points that are established in advance, enabling the augmented reality system to quickly and accurately locate and align with the target object without requiring complex real-time image processing.
Solution Approach 2:
The patent replaces complex image processing mechanisms with a simpler geometric matching approach. Instead of analyzing and processing image data to identify objects, the system directly compares the camera's view of characteristic measurements with pre-defined three-dimensional landmark coordinates, significantly reducing computational complexity while maintaining recognition accuracy.
Data Source
AI summary
A method for the recognition of real objects, in particular three-dimensional objects, for augmented reality engines comprises the steps of: producing at least one three-dimensional drawing of the real object; defining at least one characteristic measurement of the real object starting from the three-dimensional drawing; entering in an electronic device said at least one characteristic measurement of the real object; positioning the electronic device in such a way as to view the real object; detecting an image of the real object in the electronic device; superposing said at least one characteristic measurement of the real object on the image detected; aligning said at least one characteristic measurement of the real object with the image detected in such a way as to determine the positioning of the electronic device relative to the real object.