We introduce a new encryption notion called Euclidean Distance based Encryption (EDE). In this notion, a ciphertext encrypted with a vector and a threshold value can be decrypted with a private key of another vector, if and only if the Euclidean distance between these two vectors is less than or equal to the threshold value. Euclidean distance is the underlying technique in the pattern recognition and image processing community for image recognition. The primary application of this encryption notion is to enable an identity-based encryption that incorporates biometric identifiers, such as fingerprint, face, hand geometry, vein and iris. In that application, usually the input biometric will not be exactly the same during the enrollment and encryption phases. In this poster, we propose this new encryption notion and study its construction. We show how to generically and efficiently construct an EDE from an inner-product encryption (IPE) with reasonable size of private keys and ciphertexts. We also propose a new IPE scheme that is equipped with a specific characteristic to build EDE, namely the need for short private key. Our IPE scheme achieves the shortest private key compared to existing IPE schemes in the literature, where our private key is composed of two group elements only.