Image steganography is a technique of hiding secret data into a cover image and so as to prevent the intruders from accessing the secret data. The efficiency of image steganography techniques are usually evaluated based on perceptual transparency, payload capacity, security, temper resistance and computational costs. Though there has been significant improvement in related research over the decades, available steganographic techniques usually satisfy only a subset of these criteria. This paper presents a novel IWT (Integer Wavelet Transform) based steganography method using PSO (Particle Swarm Optimization) to find the optimal substitution matrix for converting secret data into their substituted forms. In the proposed method, optimal pixel adjustment process is used to improve perceptual transparency so that the obtained stego image has low distortion. The proposed method improves the security, imperceptibility, and robustness of the secret data by hiding them into the wavelet coefficients of an image. Thus, the paper provides a detailed study of the use of PSO in three different image steganographic methods based on (i) LSB (Least Significant Bit) substitution, (ii) DWT (Discrete Wavelet Transform), and (iii) IWT. Experiments are conducted using well-known benchmark images and results are comparatively analyzed. It is found that our proposed approach of PSO based IWT outperforms both PSO based LSB and PSO based DWT from the context of standard quality metrics, statistical analysis, security level estimation, payload capacity, imperceptibility etc. We have also provided a comparative overview of the existing data hiding methods including the proposed approach.

图像隐写术是一种将秘​​密数据隐藏到封面图像中，以防止入侵者访问秘密数据的技术。图像隐写技术的效率通常基于感知的透明性，有效载荷容量，安全性，抗回火性和计算成本进行评估。尽管数十年来相关研究已取得重大进步，但可用的隐写技术通常仅满足这些条件的一部分。本文提出了一种新的基于IWT（整数小波变换）的隐写方法，该方法使用PSO（粒子群优化）来找到用于将秘密数据转换为其替代形式的最佳替代矩阵。在建议的方法中，最佳像素调整过程用于提高感知透明度，从而使所获得的隐身图像具有较低的失真度。所提出的方法通过将秘密数据隐藏到图像的小波系数中，提高了秘密数据的安全性，不可感知性和鲁棒性。因此，本文提供了基于（i）LSB（最低有效位）替代，（ii）DWT（离散小波变换）和（iii）IWT的三种不同的图像隐写方法中PSO的使用的详细研究。使用众所周知的基准图像进行实验，并对结果进行比较分析。从标准质量指标，统计分析，安全级别估计，有效载荷容量，不可感知性等方面来看，我们提出的基于PSO的IWT的性能优于基于PSO的LSB和基于PSO的DWT。