Intelligent plantar image calibration method based on machine learning
A technology of machine learning and calibration methods, applied in the field of neural network models, can solve problems such as inconvenient operation and low work efficiency, and achieve the effects of convenient operation, high work efficiency, and labor cost saving
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0022] see figure 1 , the present invention provides a technical solution: a method for intelligently marking plantar images based on machine learning, comprising the following steps:
[0023] S1. Acquiring plantar images. The plantar plane images of both feet are collected by an image acquisition device. The acquired plantar images were enlarged so that the maximum width of the sole was 20 cm. The obtained plantar image is denoised, and the plantar image is grayscale processed to obtain a grayscale image of the plantar.
[0024] S2. Manually calibrate the reflection area of the plantar image. The reflection area includes: heart, liver, lung, stomach and kidney, and send the manually calibrated image data as training data to the image calibration model.
[0025] In this embodiment, the image acquisition device uses an infrared camera.
[0026] S3. The image calibration model is trained according to the training data, and then the image data in the low-certainty area is r...
Embodiment 2
[0030] This embodiment provides an intelligent calibration method for plantar images based on machine learning. Compared with Embodiment 1, the reflection areas that need to be calibrated are added, so that the marking information of the plantar reflection areas is more and more sufficient.
[0031] S1. Acquiring plantar images. The plantar plane images of both feet are collected by an image acquisition device. The acquired plantar images were enlarged so that the maximum width of the sole was 20 cm. The obtained plantar image is denoised, and the plantar image is grayscale processed to obtain a grayscale image of the plantar.
[0032] S2. Manually calibrate the reflection area of the plantar image. The reflection area includes: heart, liver, lung, stomach, kidney, spleen, small intestine, thyroid and pancreas, and send the manually calibrated image data as training data to the image calibration model.
[0033] In this embodiment, the image acquisition device uses an infra...
Embodiment 3
[0038] This embodiment provides a machine learning-based intelligent calibration system for plantar images, including an image acquisition module, an image processing module, and an intelligent calibration module.
[0039] The image acquisition module is used to acquire plantar images. The image processing module is used to amplify the collected foot image so that the maximum width of the foot is 20cm. Denoise the obtained plantar image, and perform grayscale processing on the plantar image to obtain a grayscale image of the sole, and finally output the grayscale image of the sole as training data. The intelligent calibration module is used for training according to the training data, and then retrains the image data in the low-certainty area to obtain the trained image calibration model. The image calibration model is used for intelligent calibration of the plantar image to be measured, and outputs the reflection area Calibration result.
[0040] A machine learning-based in...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More