Researeh Of Face Image Quality Evaluation Method In Face Recognition

Image quality evaluation is a classical research filed. It aims to design algorithm whichconsistent to human subject perception. In many image processing of applications areas, theimage qualityevaluation algorithms are veryimportant. As one of the important image of thequality evaluation, face image quality evaluation plays a prominent role in recognition, facialdetectionandfacetrackingsystem.At present, many scholars have been put forward a number of image quality evaluationalgorithms. This paperdeeplystudyimagequalityevaluationmethods andanalyses twotypesof evaluation methods which are subjective evaluation methods and objective evaluationmethod, and mainly represent objective evaluation method.By compare the current methods,thispaperproposesacompositivefaceimagequalityevaluationalgorithm.Theimagesizeandlocation, contrast, clarity, and bright are choose as evaluation indexes, and a Face imagequalityevaluationsystemisdevelopedbyusingVisualC++6.0andOpenCV.Theexperimentresults show that this method could express subjective feeling into evaluation indexes, andautomaticallyselectaproperpicturefromagroupofthepicture.As the imperfection of imaging system, recording equipment, a transmission mediumand processing technology, face image quality easily come down. Filtering out the highquality image is very important to enhance the performance of the face recognition systembefore face recognition. Meanwhile, Face image quality evaluation study have an importanttheoretical meaning and social values in propelling the developing the field of facerecognition,computervision.