Important improvements for decrypted image's quality. Essay

These improvements include calibration of the visual quality of the decrypted image, decreases in pixel expansion, increases in image contrast, and the development of stenography. Stenography is the hiding of secret images or other data in a public or innocent image and the secret image or data is extracted in decryption. We will briefly cover a type of stenography later in the project.
The Notional VCS Scheme
Fig. 1.1
The resultant pixel with both black subpixels produces a black pixel and one with one black and one white subpixel produces a white pixel. We are left with an image produced from two completely unrecognizable pixilations. However, in the basic scheme, we pay heavily for this security with image quality. Since each pixel of the original image is represented by two subpixels in each share, the image is expanded laterally, giving a distorted decrypted product. Much of the fine tuning of VCS has to do with minimizing loss of image quality.
Constructing the VCS Scheme
In more definite terms, we have basis matrices representing each pixel of the original image. These binary matrices are n   where n={r_1,,r_n } , these subpixels undergo an operation mimicking the effect of superimposing them with multiple transparencies. This OR operation, as opposed to an XOR operation, is demonstrated below. Clear remains clear while one black subpixel decides the value of the whole column, denoted by the last value in the column.
While visual cryptography schemes based on the XOR operation exist, valued for their image improved quality, for the purpose of this project we will focus on OR based schemes.
Taking what we've learned about the basis matrices of VSS, let's move to the real meat of visual cryptography, image quality. Quality is decided by two main metrics, pixel expansion and contrast. We have already touched on pixel expansion. Contrast is just as, if not more, important to a decrypted image's quality. One main method of contrast maximization is known as genetic algorithms, GA. The basic idea of GA is to pick and develop the best solutions out of a field of competitors using the biological evolutionary concept in which a population of p chromosomes is subjected to four different functions of GA reproduction, crossover, elitism, and mutation. In the context of this adaption of GA, chromosomes are described as, for example, a sixteen-bit binary string for a (2,2) cryptography scheme with 2 x 4 basis matrices. Reproduction selects a subset of the population P to move on to the next generation using stochastic universal sampling. Crossover takes two chromosomes from the reproduction selected chromosomes and crosses their bits to produce two new "offspring" chromosomes. Elitism merely takes the chromosomes with the greatest initial fitness values, denoted _0 which will be covered later, and pushes them along to the next generation without any additional selection. Mutation is the random selection of chromosomes out of the population for bit flipping. With these principles in mind, we move to adapt GA to visual cryptography.