Hash functions-based random number generator with image data source

In modern cryptosystems, real random numbers play critical roles in most key management systems to make sure that the generated key is able to provide high level of security in order to protect the sensitive information and assets in highly secured information systems. One of the main challenges in generating random values is the needs for special noisy source such as nuclear decay, brownian motion, analog circuit and quantum mechanics for producing the random values. Beside the cost, complexity and difficulty of using these special noisy sources as the source of random number generator, there is also the need for additional hardware devices to communicate with these sources. In this paper, we proposed an inexpensive and less complicated method to generate true random values without any need for especial noisy sources. We explain how we use hash functions in combining two data extractors to transform normal image data and make it suitable as the source of random number generator.