JPEG Metadata: A complete study
| Autor: | Nithyanandam Pandian, Jincy J. Fernandez |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Computer science
business.industry Digital forensics ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Data_CODINGANDINFORMATIONTHEORY computer.file_format Image segmentation JPEG Metadata Discrete cosine transform Computer vision Image file formats Artificial intelligence business computer Digital watermarking Transform coding |
| Zdroj: | 2018 International Conference on Recent Trends in Advance Computing (ICRTAC). |
| DOI: | 10.1109/icrtac.2018.8679322 |
| Popis: | JPEG is one of the widely used image format. As like other image formats, JPEG has its own special bytes called JPEG markers or metadata that gives relevant information about the image. In applications such as watermarking, image theft identification and digital forensics, where some kind of confidential information is embedded into the image either for copyright protection or image authentication. Usually this embedding is done on DCT coefficients. But with the use of quantization technique and compression nature of JPEG, the extraction/recovery of embedded data may not be possible all the times. A solution to this is to use quantized DCT coefficients for embedding. In order to decode the compressed data into quantized DCT coefficients, the encoding method used and also the type of chroma-subsampling used are to be known. This information is stored in the JPEG marker segment along with other relevant information. This paper detailed about the various types of JPEG markers and has shown experimental results. This will help young researchers to get the basic idea about the JPEG metadata. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|