Which of the following is not used in standard JPEG image compression ?
2016
Which of the following is not used in standard JPEG image compression ?
- A.
Huffman coding
- B.
Runlength encoding
- C.
Zig-zag scan
- D.
K-L Transform
Attempted by 314 students.
Show answer & explanation
Correct answer: D
Answer: K-L Transform is not used in standard JPEG.
Why:
JPEG uses a fixed 8×8 block Discrete Cosine Transform (DCT) to decorrelate image data before quantization.
After quantization, a zig-zag scan orders coefficients from low to high frequency to create long runs of zeros.
Run-length encoding compresses those runs of zeros, and Huffman coding performs the final entropy coding.
Why K-L Transform is not used in standard JPEG:
The K-L transform (PCA) is data-dependent: it requires computing eigenvectors for the image statistics, which is computationally intensive and would require transmitting transform information to the decoder.
The fixed DCT is a good, low-complexity approximation to the K-L transform for typical images and is separable and fast to implement, making it suitable for a standard.
Therefore, the K-L Transform is not part of standard JPEG, while Huffman coding, run-length encoding, and zig-zag scan are all used.