Image Recognition With Auxiliary-Part Spatial Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Visually impaired individuals face challenges in accurately and completely reading text information due to difficulties in capturing clear photos and maintaining spatial alignment, leading to high error rates and deviations when using optical character recognition and text-to-speech technologies.
Innovation Solution
An image recognition method that prompts users to establish a location association between an auxiliary part and a to-be-recognized object, allowing spatial alignment in three degrees of freedom, using proprioception to maintain terminal positioning and defining an area of interest through computer vision, with reminders for posture and location adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional OCR and TTS technology is used to help visually impaired persons read text information, then text recognition capability is provided, but the user cannot accurately and completely capture text due to lack of visual feedback, leading to high error rates
Solution Approach 1:
The patent introduces an auxiliary object as an intermediary between the user and the target object. The auxiliary object serves as a reference marker that the user can easily locate and align with the target using proprioception. This intermediary element bridges the gap between the user's limited spatial awareness and the precise positioning required for accurate text capture, eliminating the need for visual feedback during alignment.
Solution Approach 2:
The patent replaces the visual-mechanical alignment system with a proprioceptive-mechanical system. Instead of relying on visual feedback to guide mechanical positioning of the terminal, the system uses proprioceptive awareness of the auxiliary object's location to guide alignment. The computer vision system then detects the auxiliary object and automatically calculates positioning parameters, substituting manual visual-mechanical adjustment with automated sensor-based positioning.
2Measurement precision
If the user is guided to move the terminal in four degrees of freedom (three displacement and one turn) to align with the target object, then spatial alignment is achieved, but the user is prone to deviation and the error rate is high
Solution Approach 1:
The patent segments the alignment operation into two independent parts: (1) locating and aligning with the auxiliary object using proprioception (simplified to basic positioning), and (2) automatic terminal orientation adjustment performed by the system. This segmentation reduces the user's operational burden from controlling four degrees of freedom to simply positioning the terminal near the auxiliary object, while the system handles the complex rotational and orientational adjustments automatically.
Solution Approach 2:
The auxiliary object serves as a fixed reference intermediary that simplifies the alignment process. Instead of directly aligning with the target object which requires complex four-degree-of-freedom control, the user aligns with the auxiliary object which has a simpler, more predictable spatial relationship. The system then uses the auxiliary object's known position to calculate and execute the precise terminal orientation needed to capture the target.
3Manufacturing precision
If the user manually adjusts the terminal position and orientation to capture clear photos, then photo quality is improved, but the user cannot accurately quantify moving distance and turning angle, increasing deviation from target
Solution Approach 1:
The patent replaces manual mechanical positioning and orientation adjustment with an automated system based on computer vision and sensor data. The system uses the camera to detect the auxiliary object's position, calculates the required terminal movement and rotation parameters, and guides or automatically executes the adjustment. This substitution eliminates the user's inability to accurately quantify movement parameters, as the system performs these measurements and calculations automatically using optical and sensor-based methods.
Solution Approach 2:
The patent implements a feedback loop where the terminal continuously captures images, detects the auxiliary object and target object positions through computer vision, calculates positioning deviations, and provides guidance feedback to the user for adjustment. This closed-loop feedback system enables precise positioning by constantly monitoring and correcting the terminal's position and orientation based on detected object locations, allowing the user to achieve accurate photo capture without needing to manually estimate movement parameters.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
An image recognition method includes: outputting a first reminder, where the first reminder indicates a user to establish a location association between an auxiliary part and a to-be-recognized object, and control a terminal to photograph the auxiliary part, and when the auxiliary part exists in a shot first image and a target object whose location relationship with the auxiliary part meets a first preset condition exists in the first image, obtaining a recognition result of the target object based on a captured second image, where the first image and the second image are images in a video stream that is shot by the user controlling the terminal after the first reminder is output, and capture time of the second image is later than that of the first image. According to this application, the user is prompted to establish the location association between the auxiliary part and the to-be-recognized object. This reduces action costs of the user and increases efficiency of recognition.