Computational Intelligence in Machine Learning by Unknown

Author:Unknown
Language: eng
Format: epub
ISBN: 9789811684845
Publisher: Springer Nature Singapore

3 Design of New Work

The input for the system is the video feed from the surveillance camera, from this video feed, the frames or images are extracted. Basically, all videos are made up of frames [8]. From these images, the initial image without students which is background image is constructed. Next are the usual frames with students, now we can consider a pair of images which are the current image and the other being background image. The difference of these two images is computed and if the difference is greater than the pre-defined threshold value then we increment the counter, based on how many times, the counter was incremented the decision will be made regarding anomalous activity being detected.

The background is static or say constant but the current image is not content, it is different based on the activity of the student in the video feed which is captured in the frames. Before this process of image subtraction, all the current images are required to change their color scale. There are various color scales provided but the common color scale which works efficiently in video based applications is gray scale. The reason to choose gray scale is that we don’t really use the color attributes of the images/frames to analyze the images and detect the anomalous activity.

Initial steps involve establishing a connection to the camera from which the video feed is generated which servers as input to the system. It involves the following steps:

1.cv2.videocapture()

2.Infinite loop—read()

3.cv2.imshow()

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.