白丝美女被狂躁免费视频网站,500av导航大全精品,yw.193.cnc爆乳尤物未满,97se亚洲综合色区,аⅴ天堂中文在线网官网

Facial image processing method and apparatus, electronic device and computer readable storage medium

專利號(hào)
US11176355B2
公開(kāi)日期
2021-11-16
申請(qǐng)人
BEIJING MICROLIVE VISION TECHNOLOGY CO., LTD(CN Beijing)
發(fā)明人
Jinfeng Lai; Youwen Zhuang
IPC分類
G06K9/00; G06T7/50; G06T5/20
技術(shù)領(lǐng)域
image,face,module,in,depth,processing,feature,be,obtain,first
地域: Beijing

摘要

A facial image processing method and apparatus, an electronic device and a computer readable storage medium. The facial image processing method comprises: acquiring an image to be processed; detecting whether a facial image exists in the image to be processed; in response to the detected facial image, carrying out a first processing on the facial image to obtain a first image; and obtaining depth information, and adding the depth information to the first image to obtain a processed image. According to the method, the image depth information can be added after the image is processed, such that the image has a stereoscopic impression and the user operation is simplified.

說(shuō)明書

In step S205, a face feature point of the first image is obtained, depth information of the feature point is obtained from a preset standard template, and the depth information is mapped to the feature point of the first image, to obtain a processed image.

After the first image is obtained, the face feature point on the first image is obtained. The feature point is obtained by using a preset algorithm. The depth information of the feature point is obtained from a preset standard template. The standard template includes a standard face image and standard feature points on the face image, where the feature points include respective depth information and have respective numbers. The first image is divided in a same manner as that of the standard template, and each feature point in the first image has a corresponding feature point in the standard template. The depth information of the feature point in the standard template is mapped to the corresponding feature point in the first image, to obtain the processed image. There may be multiple standard templates, which may be face images captured at different angles, to adapt to faces of different angles in the image, so that the depth information is more accurate. In a case of multiple templates at different angles, the coordinates of the feature points are recorded. If coordinates of the feature point in the first image satisfies a predetermined relationship with the coordinates of the feature point in the standard template, the depth information of the feature point in the standard template is obtained, wherein the predetermined relationship may be that the difference between the coordinates of the feature point in the first image and the coordinates of the feature point in the standard template is within a threshold.

權(quán)利要求

1
微信群二維碼
意見(jiàn)反饋