Tire-size identification method, tire-size identification system and computer-readable storage medium
AssigneeWistron Corporation
InventorsZhe-Yu Lin, Chih-Yi Chien
Life of this patent
Estimating the term…
Abstract
A tire-size identification method includes the following steps. An object-detection model frames a tire image in an image to generate a framed tire image. The framed tire image is input into an image-segmentation model. The image-segmentation model outputs the tire inner diameter of the tire. Moreover, the tire inner diameter is input into a classification model. The classification model outputs the tire size.
Claims17
A tire-size identification method, comprising:
- inputting an image into an object-detection model, wherein the object-detection model frames a tire image in the image to generate a framed tire image;
- inputting the framed tire image into an image-segmentation model, wherein the image-segmentation model outputs a tire inner diameter of the framed tire image;
- inputting the tire inner diameter into a classification model, wherein the classification model outputs a tire size according to an inner diameter pixel value of the tire inner diameter for classifying;
- and
- after inputting the inner diameter pixel value of the tire inner diameter into the classification model, the classification model outputting a probability corresponding to each of a plurality of candidate tire-size categories, and using a tire-size category corresponding to a highest probability of the probabilities as the tire size.
A tire-size identification system, comprising:
- a processor;
- and
- a storage device, wherein the processor accesses a program stored in the storage device to implement an object-detection model, an image-segmentation model and a classification model;
- wherein the object-detection model frames a tire image in an image to generate a framed tire image;
- wherein the image-segmentation model outputs a tire inner diameter of the framed tire image;
- wherein the classification model outputs a tire size according to an inner diameter pixel value of the tire inner diameter for classifying;
- and
- wherein after the processor inputs the inner diameter pixel value of the tire inner diameter into the classification model, the classification model outputs a probability corresponding to each of a plurality of candidate tire-size categories and uses a tire-size category corresponding to a highest probability of the probabilities as the tire size.
A non-transitory computer-readable storage medium that stores one or more programs comprising instructions that, when executed by a processor, cause the processor to perform the following steps:
- inputting an image into an object-detection model, wherein the object-detection model frames a tire image in the image to generate a framed tire image;
- inputting the framed tire image into an image-segmentation model, wherein the image-segmentation model outputs a tire inner diameter of the framed tire image;
- inputting the tire inner diameter into a classification model, wherein the classification model outputs a tire size according to an inner diameter pixel value of the tire inner diameter for classifying;
- and
- after inputting the inner diameter pixel value of the tire inner diameter into the classification model, the classification model outputting a probability corresponding to each of a plurality of candidate tire-size categories, and using a tire-size category corresponding to a highest probability of the probabilities as the tire size.
Dependencies are inferred from the claim text.