A face cartoonization processing method based on computer vision

By detecting and expanding the facial area in complex environments and accurately acquiring the ears and hair, the problem of incorrect and missed recognition of key head features is solved, and the accuracy of the cartoonized results and background preservation are achieved.

CN115049763BActive Publication Date: 2025-09-05SHENZHEN MAXVISION TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210480918.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-05-05
Publication Date
2025-09-05
Estimated Expiration
2042-05-05

AI Technical Summary

Technical Problem

Existing technologies are not accurate in capturing key features of the human head in images in complex environments. Especially when multiple people are present at the same time, recognition errors, omissions, and confusion are prone to occur, resulting in cartoon results that are not in line with common sense.

Method used

By obtaining the initial image, detecting the face area and expanding it proportionally, accurately detecting the ear and hair areas, performing segmentation and fusion processing, and finally mapping the cartoonized result image back to the initial image to maintain background invariance.

Benefits of technology

Accurately capture key head features to avoid distorted cartooning results when multiple people are in the same frame, maintain background invariance, and achieve accurate extraction and cartooning of ears, hair, and faces.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115049763B_ABST
    Figure CN115049763B_ABST
Patent Text Reader

Abstract

The present application provides a computer vision-based face cartooning processing method, comprising: obtaining an initial image; detecting a face in the initial image to obtain a face area; expanding the face area proportionally to obtain an expanded area; detecting ears and hair in the expanded area to obtain an ear area and a hair area; performing ear segmentation, hair segmentation, and face segmentation operations in the expanded area, respectively, and outputting an ear segmentation image, a hair segmentation image, and a face segmentation image; fusing the ear segmentation image, the hair segmentation image, and the face segmentation image to output a head key feature fused segmentation image; cartooning the head key feature fused segmentation image to output a face cartoonization result image; mapping the face cartoonization result image back to the initial image according to coordinates, and outputting a final face cartoonization image with a background; the computer vision-based face cartooning processing method of the present application can accurately obtain the head key features.
Need to check novelty before this filing date? Find Prior Art