基于人脸图像的性别识别方法研究

摘要

Study gender identification. Most traditional gender identification methods depends on some obvious human secondary sex characteristics, when the face's gender characteristics are not obvious, the successful identification rate is not high. This paper proposed a kind of gender recognition model based on gender judge method. The camera equipment was used to acquire single - frame facial images. First, the key parts of effective face feature points were acquired, and then, the effective gender recognition model was set up using the judgment model. Experimental results show that compared with traditional gender identification methods, the parameters identification accuracy was obviously improved, which achieved good recognition effect.%研究性别优化识别问题,在当前的性别识别方法中,大部分依靠的是一些明显的人体第二性征作为识别依据,当人脸上性别特征不明显的时候,识别成功率不高.为了提高准确率,提出了一种性别模型的性别识别判定方法.利用摄像设备采集单帧人脸图像,首先采集人脸关键部位的有效特征点,建立有效的性别识别模型,运用模型进行性别的判定.实验结果表明,方法相对于以人体性别特征为参数的性别识别方法,识别的准确率有了明显改进,取得了不错的识别效果.