简体   繁体   English

执行面部识别时的opencv light问题

[英]opencv light issue while performing face recognition

I am currently working on Face recognition using opencv with LBPH algorithm. 我目前正在使用带有LBPH算法的opencv进行人脸识别。 But I am getting issue that I have trained many faces at evening of different person and it was working at that time, but the next day I am not getting the same result. 但是我遇到了一个问题,我在另一个人的晚上训练了许多面孔,并且当时正在工作,但是第二天却没有得到相同的结果。

What are the best possible ways to handle this issue? 解决此问题的最佳方法是什么?

I will really appreciate any help. 我将非常感谢您的帮助。

In order to increase your Face Recognition System accuracy you should train your network on images of the different individuals photographed in different ambients, at different time during the day, from different distances, in different poses and with different illumination setups. 为了提高人脸识别系统的准确性,您应该在一天中的不同时间,不同距离,不同姿势和不同照明设置下,在不同环境下拍摄的不同个体的图像上训练网络。

Once you have acquired this dataset, retrain your network on it and you should see that the predicition confidence increases and false positive decrease accordingly. 一旦获取了此数据集,就可以对其进行重新训练,您应该看到谓词的置信度增加了,而假阳性的减少了。

Hope this helps. 希望这可以帮助。

声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM