Intelligent System for Face Recognition Using Soft Computing Approach

Abstract

In the current trend of banking and other security systems vision based analysis plays a very important role. The Artificial intelligence helped people to build devices and machines which can make life easier and do difficult tasks easily. Using human machine interface, the security systems in banking etc were moved from password based security to image processing method. For security and image processing applications, detection of face from a camera image should be done with high efficiency since most images are frames of a video. The process becomes complex if the image is captured in a bad light or worst climatic conditions. The past methods voluntary action has the disadvantages of being tough to use as well as non adaptable for covert use as in surveillance applications. But the capturing and acquisition units used to record an image based on facial information suffers from noise components. The noise components are random occurring events happening due to environment, recording circuits, channels and thermal form. The noises occurring create a problem like salt and pepper effect, blur, speckles etc. The problem increases if the face image is captured using a low resolution camera unit. The feature extraction method should extract the maximum number of feature available in the image. On the other hand wrongly identified features may make the recognition process error. This has motivated to work in this area. The problem in extraction of facial features and recognition is taken for this research and solution was arrived. The feature extraction and recognition process with a preprocessing stage is proposed with a new methodology. newline