Abstract: The QR code is introduced into the optical image encryption system as a data container, which can remove the speckle noise generated by the optical system when the image decryption is carried out, so that the decrypted image can be restored without loss. It is a very ideal and practical method for removing the noise. However, due to the limited storage capacity of QR code, only a few letters, short sentences or a very small(32×32 pixel) grayscale image can be stored, which seriously restricts the further application of this technology. In order to store a larger grayscale image into the QR code, an image lossless compression method based on arithmetic coding was proposed. First, the grayscale image was read as the binary data in the form of a file. Then, converted it to decimal data and compressed it into the binary data by using the arithmetic coding. Finally, converted it to decimal data again. The experimental results show that this method has high compression efficiency, which can store a 64×64 pixel image with rich grayscale values into a 31-version QR code for encryption and decryption successfully.