how to change compression level in PCA?

1 visualización (últimos 30 días)
azizullah khan
azizullah khan el 21 de Feb. de 2014
Comentada: azizullah khan el 24 de Feb. de 2014
how to change compression in pca and how to find the compression ratio in pca...

Respuesta aceptada

Walter Roberson
Walter Roberson el 21 de Feb. de 2014
pca is not in itself a compression scheme. pca is sometimes used for compression by finding the vectors that are most responsible for variation in the image and discarding some of the vectors that are least responsible (which are closer to 0.) To change the compression level, discard more or fewer components.
  3 comentarios
Walter Roberson
Walter Roberson el 22 de Feb. de 2014
"intensity reduction" might be called "dynamic range compression" in some cases.
In general, "compression" is "to make smaller" and that is the sense used for dynamic range compression (amongst other uses.)
When compression of data is being discussed, the topic is compressing (making smaller) the memory required to reproduce the data afterwards ("decompression"). There are two important sub-varieties of data compression: lossless (in which decompression must be able to exactly recreate the data), and lossy (in which decompression just has to recreate the data "close enough to be usable" for whatever purpose the compression was devised.
In the more general case, "resizing" an image does not necessarily involve reducing the amount of data the image requires, and might only refer to the amount of display space used for the image.
In the case of a data array such as might be used to store an image, "resizing" it does change the amount of storage used. Resizing a data array to be smaller is one form of lossy compression.
Using PCA for image compression is a form of lossy compression: you keep the information that has the most effect on how the image looks, and you let some of the fine details be erased by not storing the information needed to recreate them.
azizullah khan
azizullah khan el 24 de Feb. de 2014
thanks..................

Iniciar sesión para comentar.

Más respuestas (0)

Categorías

Más información sobre Dimensionality Reduction and Feature Extraction en Help Center y File Exchange.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by