Image encryption is a widely used method of image information protection, and multiple-image encryption methods are rapidly evolving in the face of the exploding amount of information. Aims to solve the problem that existing optical image encryption techniques frequently experience optical contamination, which degrades the quality of the decryption image, a multi-image encryption method based on computational holography by combining Quick Response (QR) codes and Fresnel phase-only holograms is proposed. We expect to introduce a data container QR code with an error correction mechanism, where the information is transformed into a Fresnel phase-only hologram in the form of a QR code. It can be cleared even when suffering from optical contamination, within the error correction limits of the data containers, by using their error correction mechanisms and fast information acquisition capabilities. The original information is protected from optical contamination and high-quality, even lossless, results from optical decryption are produced. The encryption process of our method firstly employs a QR code encoder to convert the plaintext image to be encrypted into a corresponding dynamic QR code, and then the QR code is used as the target image to generate a Fresnel phase-only hologram corresponding to the target image using the Gerchberg-Saxton algorithm in the Fresnel domain. These phase-only holograms are transformed into integers through linear mapping; subsequently, with the help of a chaotic system, a noise-like distribution made up of chaotic sequences is generated; multi-image integration is then achieved using the exclusive OR (XOR) superposition operation, and the result of the obtained operation serves as the final ciphertext. The decryption process is the inverse of the encryption process, in which the XOR superposition operation between the ciphertext and the decryption key of the target image extracts the Fresnel phase-only hologram of the image to be decoded. Next, the pixel value of the hologram is inversely mapped to a real number. This real number is then combined with the optical key of the hologram for the purpose of reconstructing the QR code. Finally, the recovered QR code is decoded using a decoder to obtain the plaintext image. In this paper, numerical simulations and optical experiments are used to verify the effectiveness of the proposed method. First, for the numerical simulation of encryption and decryption, various sizes of color, binary, and greyscale images are chosen. Second, a phase-only Space Light Modulator (SLM) based on liquid crystal on silicon (LCoS) is loaded with the generated Fresnel phase-only holograms, and SLM propagates the modulated light waves to the imaging surface. This optical system is constructed to verify the true state of QR codes against optical contamination. The experimental results show that a decoder can extract the original image from the optical reconstruction results. The reconstructed QR code is eventually captured by a camera as the modulated light wave propagates to the image plane. On the one hand, regarding the key sensitivity, since the security of the proposed method depends on the chaotic key and the optical key, first, the chaotic key sensitivity is verified by adjusting the control parameter, initially set value, and the number of initial chaotic sequences discarded by the chaotic system, respectively. The experiments confirmed a tiny change in the three parameters will result in incorrect decryption results. Second, we carried out sensitivity simulation experiments on optical keys by altering only the reconstruction distance and wavelength. Similarly, we discovered that a small parameter change resulted in incorrect decryption results, indicating that the key is sufficiently sensitive. Analyzing the robustness is crucial since, on the other hand, there is always a chance of information loss during the sharing of information. In this study, a series of cropping attack tests have been developed to simulate varying degrees of data loss to thoroughly assess the anti-data loss efficacy of the encryption method in question. We used three different data loss scenarios in our study: 3.5%, 16%, and 25%. The experimental results demonstrate that, despite the various degrees of data loss, the decrypted QR code image is still able to decode the original image. Noise is unavoidable during data transmission, both salt & pepper noise and Gaussian noise are analyzed. The first case, the salt & pepper noise with intensities of 5%, 10%, and 20% is added to the ciphertext, respectively. The second case, the Gaussian noise with intensities of 0.5, 0.8 and 1.0 (means: 0, sigma: 0.05) is added to the ciphertext, respectively. All of reconstructions can be decoded to produce the original image, and the experimental findings demonstrate that they are still within an acceptable range. This shows that the proposed method has robustness in complex environments and its ability to sustain more steady decryption results in the face of numerous unfavorable factors. Therefore, this method offers a highly practical and reliable alternative path for multi-image encryption in the field of optical information security. With the application of this method, which provides multi-image encryption without the use of optical dimension multiplexing techniques, the difficulty of developing encryption systems is greatly reduced. A redundant error correction system is conceived to protect the target image from optical pollution by marrying holograms with QR codes.