A new coding initiative is introduced “Visually lossless” coding in place of numerically lossless coding to reduce the piling space and lower data transmission. The question for introducing this new coding method arises due to the increased image sizes. Here we speak about the lossy compression method on encrypted image, which compresses the image with minute quality losses that cannot be detected. This lossy compression method includes DCT with RLEcompression, Hierarchical Oriented Prediction (HOP), uniform quantization, orthogonal matrix generation, negative sign removal and Huffman compression. The encrypted image is divided as elastic part, which is compressed using Xingpeng Zhang method and rigid part for which HOP method is used for compression. This method is tested in different type and size of images and the results are obtained. The results reveal that this method is much better than the existing compression methods. The bit rate reduction ratio of this method is 10.45% and the naked eye perception is visually lossless.