Abstract— This paper proposes a new steganographic method – based on Roberts operator edge detector in digital images – with the aim to increase the size of secret messages that can be embedded and hidden in images. An algorithm is developed to execute the task with two constraints: i) fixed dimensions size of the image and ii) no visual distortion of the resulted image. The proposed methodology embeds data of the built-in secret message into the image’s boundary pixels obtained using the Roberts operator. To do this, edge detector threshold selection strategy is proposed to achieve the optimal threshold value by creating a reference images with different numbers of boundary pixels. Finding this optimal threshold value is very important since it decreases the image’s visual distortion. The results of the conducted experiments to test the developed algorithm show excellent results in hiding secret messages in the low-order bits of the selected pixel codes of the reference image boundaries.