Laser speckle has been planned in a very variety of works as a high-entropy supply of unpredictable bits to be used in security applications. Bit strings derived from speckle may be used for a range of security functions like identification, authentication, anti-counterfeiting, secure key storage, random variety generation and tamper protection. The selection of laser speckle as a supply of random keys is kind of natural, given the chaotic properties of speckle. This work proposes an optical device speckle recognition system for credibility verification. Owing to the distinctive state surfaces of objects, laser speckle provides diagnosable features for authentication. A Gabor filter, SIFT (Scale-Invariant Feature Transform), and projection were wont to extract the features of optical device speckle images. To accelerate the matching method, the extracted Gabor features were organized into an indexing structure using the modified K-means algorithm. The special relations among the matching points are then remodeled to 9DLT (Direction Lower Triangular) representations. Then, the Frequent Pattern Growth (FP-Growth) algorithm mines frequent patterns therefore a helpful association rule is obtained because the feature to spot the similarity between every of the speckle pictures for the aim of credibility verification. Plastic cards were used because the target objects within the planned system and also the hardware of the speckle capturing system was designed. The experimental results showed that the retrieval performance of the planned methodology is correct once the information contains 516 optical device speckle pictures. The planned system is powerful and possible for credibility verification.

Image Processing, Imaging Systems, Pattern Recognition, Optical Security and Encryption, K-means algorithm, FP-Growth algorithm, Direction Lower Triangular.