基于稀疏表示的人脸图像智能抠图方法  

Intelligent Matting Method of Face Image Based on Sparse Representation

作　　者：赵敏 ZHAO Min(Puyang Vocational and Technical College,Puyang Henan 457001,China)

机构地区：[1]濮阳技师学院,河南濮阳日457001

出　　处：《信息与电脑》2023年第7期26-28,共3页Information & Computer

摘　　要：针对现有方法后期抠图后与新图像连接度低的问题,提出基于稀疏表示的人脸图像智能抠图方法。首先设计人脸图像变换模块,其次构建稀疏表示人脸图像智能抠图模型,最后采用调整掩膜透明度的方式强化智能抠图的效果。测试结果表明,该方法的抠图效果更佳,连接紧密度较高,具有较高的应用价值。Aiming at the problem of low connectivity between the existing methods and the new image after matting,an intelligent matting method of face image based on sparse representation is proposed.Firstly,a facial image transformation module is designed.Secondly,an intelligent matting model for sparse representation of facial images is constructed.Finally,the effect of intelligent matting is enhanced by adjusting the transparency of the mask.The test results show that this method has better matting effect,higher connection tightness,and high application value.

关 键 词：稀疏表示 人脸图像 智能抠图 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]