A method for monitoring attendance of a teacher in class, wherein, an image capture device is installed in a classroom, an image capture area of the image capture device comprises a platform area of the classroom, and any one of class hours corresponding to the classroom is divided into a plurality of time periods, the method comprises:
- acquiring an image of the platform area captured by the image capture device installed in the classroom during each time period of one class hour;
- determining whether a face image is contained in the image;
- if a face image is contained in the image, acquiring at least one picture by the image capture device;
- and determining, for each picture, an attendance result of the teacher corresponding to the classroom during this time period based on a similarity between a to-be-recognized face image in this picture and an image in a face comparison database;
- determining an attendance result of the teacher corresponding to this classroom during this class hour based on the attendance result of the teacher corresponding to this classroom during each of the time periods,
- wherein, determining, for each picture, an attendance result of the teacher corresponding to the classroom during this time period based on a similarity between a to-be-recognized face image in this picture and an image in a face comparison database comprises:
- acquiring, from the face comparison database, the image of the teacher who is course-scheduled in this classroom during this class hour;
- determining, for each picture, whether the similarity between the to-be-recognized face image in the picture and the acquired image is greater than a predetermined similarity threshold;
- and if the similarity between the to-be-recognized face image in the picture and the acquired image is greater than the predetermined similarity threshold, increasing the number of recorded pictures by 1;
- determining whether the number of the recorded pictures is greater than a set number threshold;
- if the number of the recorded pictures is greater than the set number threshold, determining that the attendance of the teacher corresponding to this classroom during this time period is normal;
- otherwise, determining that the attendance of the teacher corresponding to this classroom during this time period is abnormal,
- or
- determining, for each picture, an attendance result of the teacher corresponding to the classroom during this time period based on a similarity between a to-be-recognized face image in this picture and an image in a face comparison database comprises:
- acquiring, from the face comparison database, the image of the teacher who is course-scheduled in this classroom during this class hour;
- determining the similarity between the to-be-recognized face image in each picture and the acquired image, and determining a maximum value of the similarity;
- determining whether the maximum value of the similarity is greater than a predetermined similarity threshold;
- and if the maximum value of the similarity is greater than the predetermined similarity threshold, determining that the attendance of the teacher corresponding to this classroom during this time period is normal;
- otherwise, determining that the attendance of the teacher corresponding to this classroom during this time period is abnormal,
- wherein, the method further comprises:
- decreasing the similarity threshold when it is determined that the attendance of the teacher corresponding to this classroom during this time period or this class hour is abnormal, and the attendance result of the teacher corresponding to this classroom during this time period or this class hour is corrected;
- increasing the similarity threshold when it is determined that the attendance of the teacher corresponding to this classroom during this time period or this class hour is normal, and the attendance result of the teacher corresponding to this classroom during this time period or this class hour is not corrected, or the similarity between the to-be-recognized face image in each picture and the acquired image is greater than the similarity threshold, and
- wherein, decreasing the similarity threshold comprises:
- decreasing the similarity threshold based on a set first proportionality coefficient;
- determining whether the decreased similarity threshold is lower than a preset minimum threshold;
- and if the decreased similarity threshold is lower than the preset minimum threshold, adjusting the similarity threshold to the minimum threshold;
- or
- determining a minimum value of the similarity between the to-be-recognized face image in each of the pictures and the image of the teacher who is course-scheduled in this classroom during this class hour;
- determining whether the minimum value is lower than the preset minimum threshold, and if the minimum value is not lower than the preset minimum threshold, adjusting the similarity threshold to the minimum value;
- increasing the similarity threshold, comprises:
- increasing the similarity threshold based on a set second proportionality coefficient
- determining whether the increased similarity threshold is greater than a preset maximum threshold;
- and if the increased similarity threshold is greater than the preset maximum threshold, adjusting the similarity threshold to the maximum threshold;
- or
- determining the minimum value of the similarity between the to-be-recognized face image in each of the pictures and the image of the teacher who is course-scheduled in this classroom during this class hour;
- determining whether the minimum value is higher than the preset maximum threshold, and if the minimum value is higher than the preset maximum threshold, adjusting the similarity threshold to the minimum value.