A novel bit-plane chaotic image encryption algorithm is proposed further to improve image encryption systems' security and anti-attack ability. This paper combines bit layer scrambling and Deoxyribonucleic Acid (DNA) encoding. In this paper, we design a novel bit-plane scrambling method using parity alignment and plaintext correlation chaotic sequences, which can significantly reduce the correlation between image pixels and the correlation between RGB layers. Then the intermediate ciphertext is DNA encoded by hyper-chaotic sequences with plaintext correlation, after the designed multiple operation diffusion, the ultimate ciphertext is generated. Furthermore, devised a new way of generating keys, which are highly correlated with plaintexts. This increases the algorithm's resistance to selective plaintext attacks. According to the simulation results, this paper's proposed image encryption algorithm can effectively resist differential and statistical attacks, has ample key space and sensitivity, and has ideal information entropy.