The face recognition attendance machine adopts the high-precision technology of today's international technology field-face recognition technology (incorporating the computer image processing technology and the principle of biostatistics), and uses computer image processing technology to extract portrait feature points from video. The principle of biostatistics is analyzed and a facial feature template is created. When the registered person walks in front of the face recognition machine, it will have a voice prompt "Hello" or the name of the person indicating that the attendance has been successful. Moreover, it also has an image update function. If the image being taken is taken as the first face, the second face is stored, and if the first face image is consistent with the second face image, the face recognition attendance machine The first face image will be automatically saved to update the second face image. The method can maintain the update of the user's face image, reduce the influence of the face shape change on the recognition, and increase the accuracy of the recognition.