You are on page 1of 2

Lossy Image Compression

Reyes Espinoza Kevin


Ingeniera en computacin Universidad Tecnolgica de la Mixteca Huajuapan de Len, Oaxaca, Mxico kenin4@gmail.com
A lossy compression is a data encoding method that compress data by discarding some of it. The objective of this method is to reduce the amount of data needed to be held o transmitted. Lossy compression is commonly used to compress media files like audio, video and images. It is needed and optimal when we can discard some of the information without noticing it or without that affecting too much to the output. Lossy Compression has lower (the lower, the better) compression ratio than lossless compression bot compressed information cannot be recovered..

II. PROCEDURE
In this practice, MATLAB environment was used to do the procedure. We used gray-scale images with 8 bits per pixel. First of all, the image was splitted into 8x8 pixels sub-images. Each sub-image was processed with a bi-dimensional DCT, then, zeroes were added to the coefficient matrix of the DCT and the inverse bi-dimensional DCT was applied next. We have to put each sub-image into an output image, in the exact place where it was taken. 3 images had to be processed. The number of zeroes that should be included in each of them is 10, 20, 30, 40, 50 and 63. For each recovered image, we rated with G for Good R for Regular or B for Bad in function of the conserved quality from the original image.

III. RESULTS
Image/Number zeroes GRIS.BMP LENA.BMP PAISAJE.BMP of 10 G G G 20 G G G 30 G G G 40 G G R 50 G R R 63 G B B Original size 256 x 256 512 x 512 256 x 256

I. INTRODUCTION
This practice goal was to verify that we can discard elements from the Discrete Cosine Transform (DCT) from an image and recover it with losses that can be barely noticed and some also pretty evident. JPEG compression uses the Discrete Cosine Transform to encode and compress data. The first part of the JPEG encode consists in Splitting the image array into 8x8 sub matrixes. Before computing the DCT of the 88 block, its values are shifted from a positive range to one centered around zero. For an 8-bit image, each entry in the original block falls in the range [0, 255]. The midpoint of the range (in this case 128) is subtracted from each entry to produce a data range that is centered around zero, so that the modified range is [-128, 127]. This step reduces the dynamic range requirements in the DCT processing. The human eye is good at seeing small differences in brightness over a relatively large area, but not so good at distinguishing the exact strength of a high frequency brightness variation. This allows one to greatly reduce the amount of information in the high frequency components. This is done by simply dividing each component in the frequency domain by a constant for that component, and then rounding to the nearest integer. This rounding operation is the only lossy operation in the whole process. This process reduces the quality of an image, and this reduction is accumulative. This means that if we compress and decompress an image, quality will be lost, but if we compress an already compressed image, we will get an even higher loss of quality. Each successive compression will cause additional quality loss.

LENA.BMP compressed using 63 zeroes.

PAISAJE.BMP compressed using 63 zeroes.

GRIS.BMP compressed using 63 zeroes.

IV. CONCLUSIONS
With the obtained results, we can notice that bigger images compressed with 63 zeroes looks clearer than smaller images compressed with the same number of zeroes. Just like we see in the compression of LENA.BMP, which is bigger than PAISAJE.BMP. Another interesting thing is that GRIS.bmp was apparently unaffected by compression. This is because of the simplicity and colors of the image. It is clear that compressed images, besides of the definition loss, has lower brightness than the original files. This is because the compression uses the variation of brightness to reduce the amount of information in the high frequency components.

REFERENCES
[1] [2] [3] Lossy Compression, WEB, revision: 31/11/13, Available http://en.wikipedia.org/wiki/Lossy_compression JPEG, WEB, revision: 31/11/13, Available in http://en.wikipedia.org/wiki/JPEG Discrete Cosine Transform, WEB, revision: 31/11/13, Available http://en.wikipedia.org/wiki/Discrete_cosine_transform in

in

You might also like