Expressions of face are an important factor in the modern human interacting systems that find the difference between the human expressive states at different times. Scrambling or encryption of data in a captured image gives a solution to an enhanced system. Consider the Internet of Things (IoT) directed distribution of image/video, the encryption of face is considered as a major part for protection of privacy. For this, biometric verification is required to get the encrypted data. After the encryption procedure the face models become chaotic signals. Using various face recognition methods, we can obtain the data, which is a traditional method. This survey paper presents the different techniques which is used to getting the face from the decrypted data by using many kernel methods.