In the current era, when everyone has adopted the Internet as an essential part of their daily lives, Internet users have projected themselves as the victim of cyber attacks. The usage of the Internet has become not only an innovation but has also led to many problems such as attacks like adware, trojan, etc. Phishing attacks have become one of the most commonly used attacks and one which is what most people are falling prey to. In this paper, we are going to present a novel technique to detect phishing attack vulnerabilities while surfing on the Internet. The proposed technique does not require any installation on the user side and it is an easy plug-and-play type of utility that can be helpful to any layman doing Internet Surfing. We have used the technologies of image processing, machine learning, and Tesseract OCR to create an alert if a URL points to a phishing website. The response time for getting this alert is comparatively lower than the maximum threshold value.