Codificação de Imagem Plenotica para Armazenamento em DNA Sintético

The main objective of this project is to propose an encoding solution for Synthetic DNA Data Storage adapted to plenoptic data. The proposed solution must be efficient, near visually lossless, and robust to errors introduced in the oligos of the FASTA file. For that the following objectives/task should be accomplished: 1. Adapt the actual coding solutions, V-DNA and HiDNA, for the coding of plenoptic data. 2. DNA data storage is a very error-prone process. The different components of the biochemical process for DNA data storage, especially sequencing, generate a lot of errors. The error rates of these processes depend on the different technologies adopted by each sequencing machine but also largely depend on the DNA code that should be embedded in a molecule. Some patterns, and other characteristics of the DNA codes have been identified as error-generating and should be avoided to make the entire data storage process functional and more reliable. Verify that the coding solutions adapted in 1 verify the identified restrictions. 3. The coding solutions evaluated in a. and b. can be divided in separate or joint source and channel coding. a. Evaluate how error-correcting codes can allow the V-DNA decoder to reconstruct the input image even when errors are introduced in the oligos of the FASTA file. b. Evaluate how error-correcting codes can allow the HiDNA decoder to reconstruct the input image even when errors are introduced in the oligos of the FASTA file. c. Compare the error correction capabilities of both schemes. The relation bitrate/distortion/error correction capacity should be study and compared. d. Compare and verify if plenoptic solutions are more error prone than standard solutions. Knowing that plenoptic image, namely holography, are known for the extra redundancy, they should be by construction less sensitive to the errors introduced in the oligos of the FASTA file. 4. For the best coding solution, develop a mechanism to allow for robustness against errors in the DNA channel. Recent experiments on actual coding solutions revealed they are not yet robust to errors in the DNA channel. An error correction mechanism should be developed to allow the decoder to reconstruct the input image even when errors are introduced in the oligos of the FASTA file. This can pass to the development of error correction quaternary codes or to the use of joint source channel coding schemes like the ones presented in recent literature. This should also consider that although the 3D contents of a plenoptic image can be very rich, the combination of several 2D views or depth slices is still not fully developed. In fact, the development of appropriate reconstruction and segmentation techniques for extended focus imaging would enable higher quality reconstructions that would take full advantage of the encoded 3D information. 5. Both the encoders proposed for JPEG DNA are trying to optimize their performance, by reflecting in their coding schemes the steps used in JPEG AI that already prove their effectiveness. Due to specific nature of plenoptic image their characteristics should be considered when implementing these optimizations. 6. Define the performance metrics for quality assessment that can be used to reliably evaluate the decoded images obtained from image coding solutions proposed in previous steps. 7. Define the subjective evaluation procedure to perceptually evaluate all decoded images quality, obtained from previous coding solutions.