Automatic Website Content Change Detection and Notification Using Image Processing

Abstract

This research develops a web application called WatchNSend that allows users to automatically monitor websites and detects changes in website content using image processing techniques. The app is built in three modules.  First, in the Setup Module the user creates a monitoring job and specifies how they would like their website monitored.  An Area Of Interest (AOI) is selected by the user within the displayed webpage, and other job details are also set, such as the desired monitoring frequency.  The app then saves the AOI image and job details for later use. The second module is the Job Module. This is where all job specifications and all job update information are stored. Also in this module, a timer controls mechanisms that monitor the specified update intervals of all jobs and initiate the third module accordingly.  The third module is the Comparison Module.  Here the app automatically collects a current copy of the job’s AOI and then compares this new version of the AOI with the previous version, using two methods called the Edge Calculation and the Overlay Calculation.  If content changes inside the AOI have reached a threshold set by the user, the app automatically notifies the user via e-mail.  Results from multiple evaluations show that WatchNSend can monitor websites, detect and analyze changes, and notify the webmaster of the changes accurately and efficiently. WatchNSend is a reliable, robust, and easy-to-use tool that can save users time while keeping them current on website changes.