- Research Article
- 10.12732/ijam.v38i9s.814
A HYBRID OF CHAOTIC MAPS WITH LYREBIRD OPTIMIZATION ALGORITHM FOR ENHANCING THE IMAGE ENCRYPTION MODEL
- Nov 03, 2025
- International Journal of Applied Mathematics
- Gagandeep Kaur,
The main purpose of the proposed image encryption (IE) method is to enhance various security characteristics to protect sensitive images on the internet. In the literature, chaotic maps are preferred for image encryption because of their high randomness and sensitivity to initial parameter values. Therefore, metaheuristic algorithms have been utilized to fine-tune the parameters. In this research, we have employed the metaheuristic Lyrebird optimization (LO) algorithm to fine-tune the chaotic map parameters. The LO algorithm is chosen over other algorithms due to its better exploration and exploitation rate to search for the best solution. The LO algorithm searches for the best parameter value based on the objective function. This research involves the formulation of a multi-objective function that incorporates numerous security parameters, namely, entropy and correlation coefficient (CC), to improve the security attributes of the IE model. The proposed IE model has two steps. In the first step, a random key is generated to transform the pixels of the secret image, whereas in the second step, the image pixels are shuffled based on the shuffling index values to reduce the correlation among the neighboring pixels. In both steps, we have used the chaotic logistic map. The proposed IE is evaluated against statistical and differential attacks on the standard dataset by considering the various images and their different resolutions. The uniform distribution of the histogram of the encrypted images is validated using the chi-square test. Besides that, entropy analysis shows that the average entropy value is 7.9973, 7.9993, and 7.9998 for 256x256, 512x512, and 1024x1024 image resolutions. The average CC and SSIM are near zero value. The average NPCR value (>0.9969) is near the 1 value. Finally, the comparative analysis based on entropy, CC, and NPCR parameters shows that the proposed IE model outperforms the existing models.
Read more