Back to results
US 12,450,801 B2Granted
Text recognition method and apparatus based on hand interaction for AR glasses
AssigneeElectronics and Telecommunications Research Institute
InventorsJong-Bae Lee, Jeung-Chul Park, Wook-Ho Son, Beom-Ryeol Lee, Yong-Ho Lee
Life of this patent
Estimating the term…
Abstract
Disclosed herein are a text recognition method and apparatus based on hand interaction for AR glasses. The text recognition method based on hand interaction for AR glasses includes collecting RGB images, extracting hand joint information from the RGB images, generating a text image based on the hand joint information, recognizing text from the text image, and outputting the recognized text.
Claims16
A text recognition method, comprising:
- collecting RGB images;
- extracting hand joint information from the RGB images;
- generating a text image based on the hand joint information;
- recognizing text from the text image:
- outputting the recognized text,
- wherein recognizing the text comprises:
- preprocessing the text image;
- and
- inferring the text by inputting the preprocessed text image to a first machine-learning model,
- wherein preprocessing the text image comprises:
- converting the text image into a grayscale image;
- adjusting a resolution of the grayscale image;
- normalizing the image of which the resolution is adjusted so that pixel values of the image are 0 to 1;
- and
- converting the normalized image into a tensor.
12 more elements
A text recognition apparatus, comprising:
- a memory configured to store a control program for text recognition;
- and
- a processor configured to execute the control program stored in the memory,
- wherein the processor is configured to collect RGB images, extract hand joint information from the RGB images, generate a text image based on the hand joint information, recognize text from the text image, and output the recognized text, wherein the processor is configured to preprocess the text image, and infer the text by inputting the preprocessed text image to a first machine-learning model, wherein the processor is configured to convert the text image into a grayscale image, adjust a resolution of the grayscale image, normalize the image of which the resolution is adjusted so that pixel values of the image are 0 to 1, and convert the normalized image into a tensor.
1 more element
Dependencies are inferred from the claim text.