A Reliable Transmission and Extraction of Textual Information Using Keyless Encryption, Steganography, and Deep Algorithm with Cuckoo Optimization

The entire work postulates a fusion of an innovative keyless image encoding approach based on a chaotic map and steganography. Nearly every single picture element has been encrypted by rearranging pixel values that are then assessed by such an adapted cat map as well as steganography has been used in the suggested methodology to transmit keyless encoded textual information utilizing a cover image, with encoded textual image information inserted into cover images. But on the other hand, retrieving textual images embedded in the receiver has indeed been regarded as the most challenging process. Furthermore, a sender will indeed transfer reasoning rather than a key to the recipient, ignoring the fact that acquiring engrained textual information hidden inside the cover image to enhance efficiency would have been challenging. Steganography was easily accomplished by placing data bits of secret textual image information into the cover image’s least significant bit of picture elements. The image quality, on the other hand, suffers greatly from the inclusion of an encrypted message. Images could indeed cause some disturbances from time to time. As a consequence, non-smooth areas of the image have been regarded as smooth. Therefore, the weighted Naive Bayes algorithm (NBA) method is being used in the illustration to retrieve the appropriate textual image information from complicated deteriorated images. Normally, images comprise a small amount of noise, so the introduction of the guided filter (GF) during the initial preprocessing phase is required. A very important feature in the textual image information extraction procedure was obtained utilizing a variety of techniques, including Gabor transform (GT), weighted Naive Bayes algorithms (WNBA), and a high-quality learning procedure for standard procedure. Eventually, textual content identification and recognition are performed using a combination of DNN and cuckoo algorithms. Again, for the proposed system, the set of data IIIT5K is being used, as well as high performance has been gained with parameters such as mean square error and peak signal-to-noise ratio.