Goal: Analyze images taken in burst mode and return the best image. Compare the images based on two parameters -Blur detection -Face and Eye detection
Platforms and Programming models used: 1.Android application 2.GPU – Grayscalling an image 3.Multithreading – Multiple threads for Image blur detection and eye detection 4.OpenCV4AndroidSDK 5.HaarCascade classifiers for face and eye detection
Limitations: 1.Our approach does not work for all rotations. Sometimes if faces are tilted or turned the algorithm may fail to detect them. 2.Eye Detections is not accurate.