Abstract— In this paper, a plaintext-related image encryption algorithm is proposed based on a newly constructed tent chaotic map in two dimensions (2D-TentCM). The proposed map extends the classical one-dimensional tent dynamics to two coupled dimensions resulting in stronger nonlinearity, improved ergodicity, and increased sensitivity to initial conditions and control parameters. To improve the plaintext dependence and to increase the effective key space, the initial states and the map parameters are adaptively updated by an image-derived key-mixing mechanism before the generation of chaotic sequences. The resulting keystreams are then used in a permutation-diffusion framework for encrypting the image data. The security analysis includes histogram uniformity, information entropy, adjacent pixel correlation, NPCR, UACI and key sensitivity analysis. The experimental results show that the proposed algorithm results in ciphertexts with near uniform statistical characteristics, strong avalanche behavior and high resistance to brute force, differential and plaintext related attacks. These results indicate that the proposed 2D-TentCM-based architecture is a competitive and computationally efficient candidate for secure image encryption applications.