used in the proposed to improve the performance of Huffman coding. The performance of
Huffman encoding depends upon unique symbol present in the secret image. Here symbol refers
to various unique pixel intensity or grayscale value. Table1 shows the experimental results of
Method 3 of [12] (R.Amirtharajan et al., 2010). There is a significant improvement in PSNR and
embedding capacity through the proposed method. The improvement obtained for LSB and
matrix embedding is exhibited in both Table2. The maximum embedding capacity of [12] and the
proposed method are shown in Table3 and Table4. The secret image constructed in the
destination is lossless since no quantization process is carried out. PSNR above 30dB is
acceptable [23] but high quality recovering (secret) image should strive for 40dB and above.
The last 64 pixel in cover image is reserved for storing the technical details, which will be used in
the receiver side to extract the secret data from the stego image. This 64 pixel (64x3=192 bytes)
should be excluded while computing the maximum hiding capacity of cover image. Secret data of
more unique symbol with any size can be hidden through our proposed method, provided it meets
the above said condition.