Hi tojeena, I suggest huffman encoding is better than the other two unfortunately if you have to encode a long run of symbol then obviously you have to choose run length encoding technique...
Hi tojeena i will say some disadvantages of arithmetic coding
1)whole codeword must be received to start decoding the symbols.
2) if a bit is corrupt in code word means the whole message will be corrupted.
3) the precision in the limit will limit to the number of encoding in an code word.
if above mentioned are not constrained to you then arithmetic code can be used...
Hope this may be helpful....
why cant you try with some other compression algorithm...
if you are interest you just send me the base paper of your project to me and i can suggest you some algorithm if possible...
For easy going you first develop the compression process of your base paper in simulink model so that you can adjust the parameter easily and find the performance of the different types of algorithm and coding techniques...instead of writing a code for long period...
then, You split the proposed algorithm in to different phase and write a code so that you can easily get through the error.... NOTE: Before choosing the algorithm you should be very clear with the CCSDS Requirements of Image compression
all the best....
Hope this may be helpful to you...