Growing requirements for preservation as well as transportation of multi-media data have been a component of everyday routine throughout the last numerous decades. Multimedia data such as images and videos play a major role in creating an immersive experience. Data and information must be transmitted quickly and safely in today’s technologically advanced society, yet valuable data must be protected by unauthorised people. Throughout such work, a covert communication as well as textual data extraction approach relying on steganography and image compression is constructed by utilising a deep neural network. Using spatial steganography, the initial input textual image and cover image are all first pre-processed, and afterwards the covert text-based images are further separated and implanted into the least meaningful bit of the cover image picture element. Thereafter, stego- images are compressed to create an elevated quality image and to save storage capacity at the sender’s end. After all this, the receiver will receive this stego-image through a communication channel. Subsequently, steganography and compression are reversed at the receiver’s end. This work has a multitude of problems that make it a fascinating subject to embark on. Selecting the correct steganography and image compression method is by far the most important part of this work. The suggested method, which integrates both image-steganography and compaction, achieves better efficacy in relation to peak signal-to-noise.