With the increasing popularity of wireless sensor networks in the IoT and Industry 4.0 era, the security of networks is critical in transmitting data and information. To address this need, we propose an optimized method for hiding and extracting information from image data. For this method, we created algorithms for hiding and extracting information based on the histogram shift method. The algorithms were developed using chessboard- and column-type prediction methods. Five different prediction methods were tested in the development of the algorithms, and the test results showed that the chessboard-type methods yielded better results with images. Then, the optimized prediction method was tested for various images along with previous methods. The results show that the proposed method has better results in terms of bits per pixel and peak signal-to-noise ratio. The method can be applied to image transmission through a wireless sensor network and provides a basis for the development of further applications.