A novel technique for random number generation using CMOS image sensor

Random Number Generators (RNGs) have been used in several traditional fields such as simulation, gaming, cryptography, etc. Random numbers are used in cryptography to generate passwords. The strength of a password depends not just on its length and complexity but mainly on its unpredictability. Strong passwords lower the overall risk of a security breach. Hence, we propose a True Random Number Generation (TRNG) using images from a Complementary Metal Oxide Semiconductor (CMOS) image sensor to provide unpredictability to the passwords. The proposed system is based on the dark pixel values of the image from the CMOS image sensor. Dark pixels are used to eliminate the requirement of light. The Image Fusion Algorithm is used in this system to combine more image files into a single image file. The process of combining more image files into one image file increases unpredictability. The TRNG does not require any additional resources, such as a light source for providing true randomness. The images received from the CMOS image sensor are processed and dark pixel values are combined to produce the initial seed value. A pseudo-random number generator algorithm is used to enhance the randomness of the initial seed value. This TRNG mobile application can be used in any android mobile phone and this can also be used in any electronic device that has a CMOS image sensor with it. The effectiveness of the system can be analyzed by statistical analysis suite.