In this era of rapid information exchange in public networks, there is a risk to information security. Data hiding is an important technique for privacy protection. Image interpolation is an important data-hiding technique in image processing. This study proposed a method called neighbor mean interpolation by neighboring pixels (NMINP) that calculates a cover image pixel by neighbor mean interpolation and neighboring pixels. To reduce image distortion, NMINP limits the number of bits when embedding secret data, making NMINP have a higher hiding capacity and peak signal-to-noise ratio (PSNR) than other methods. Furthermore, in some cases, the secret data are flipped, and the flipped data are treated in ones' complement format. A location map is not needed in the proposed method. Experimental results comparing NMINP with other state-of-the-art methods show that NMINP improves the hiding capacity by more than 20% and PSNR by 8%.