Help for Interactive JPEG Image Compression

WARNING: This is a Java program running in its interpretive mode. Be prepared to wait for a few seconds after each submission.

1. QM1 to QM6 are several predefined Quantization Matrices. QM1 is from the book "Image and Video Compression Standards" by V. Bhaskaran and K. Konstantinides, Kluwer, 1995.

Choose one of the matrices, followed by "submit" to see the JPEG encoding and decoding results.

The first row shows the encoder. The original image is at the left, the YIQ images are shown in the middle, the DCT results are at the right.
Use right mouse button to click anywhere in one of the YIQ images, the numeric values of the DCT and quantized 8x8 blocks will be displayed in the second row.
The last row shows the decoding results. The quantized DCT images are at the left, followed by the decoded YIQ, and then the decoded color image.
The compression ratio (size_of_original_image / size_of_compressed_image) is shown in the area below the quantization table.

2. In the area below the quantization table, you may type in one of the following commands:

inner operation :
- ia number / IA number: all the elements in the matrix whose x-coordinates and y-coordinates are both less than or equal to those of the current position pointed by the little rectangle are added the number, but the sum should not be more than 300.
- is number / IS number: all the elements in the matrix whose x-coordinates and y-coordinates are both less than or equal to those of the current position pointed by the little rectangle are deducted by the number, but the results should not be less than 1.
- it number / IT number: all the elements in the matrix whose x-coordinates and y-coordinates are both less than or equal to those of the current position pointed by the little rectangle are multiplied by the number, but the results should not be more than 300.
- id number / ID number: all the elements in the matrix whose x-coordinates and y-coordinates are both less than or equal to those of the current position pointed by the little rectangle are divided by the number, but the results should not be less than 1.
outer operation :
- oa number / OA number: all the elements in the matrix whose either x-coordinates or y-coordinates is more than that of the current position pointed by the little rectangle are added the number, but the sum should not be more than 300.
- os number / OS number: all the elements in the matrix whose either x-coordinates or y-coordinates is more than that of the current position pointed by the little rectangle are deducted by the number, but the results should not be less than 1.
- ot number / OT number: all the elements in the matrix whose either x-coordinates or y-coordinates is more than that of the current position pointed by the little rectangle are multiplied by the number, but the results should not be more than 300.
- od number / OD number: all the elements in the matrix whose either x-coordinates or y-coordinates is more than that of the current position pointed by the little rectangle are divided by the number, but the results should not be less than 1.
line operation :
- la number / LA number: all the elements in the matrix whose either x-coordinates or y-coordinates is equal to that of the current position pointed by the little rectangle are added the number, but the sum should not be more than 300.
- ls number / LS number: all the elements in the matrix whose either x-coordinates or y-coordinates is equal to that of the current position pointed by the little rectangle are deducted by the number, but the results should not be less than 1.
- lt number / LT number: all the elements in the matrix whose either x-coordinates or y-coordinates is equal to that of the current position pointed by the little rectangle are multiplied by the number, but the results should not be more than 300.
- ld number / LD number: all the elements in the matrix whose either x-coordinates or y-coordinates is equal to that of the current position pointed by the little rectangle are divided by the number, but the results should not be less than 1.
global operation :
- ga number / GA number: all the elements in the matrix are added the number, but the sum should not be more than 300.
- gs number / GS number: all the elements in the matrix are deducted by the number, but the results should not be less than 1.
- gt number / GT number: all the elements in the matrix are multiplied by the number, but the results should not be more than 300.
- gd number / GD number: all the elements in the matrix are divided by the number, but the results should not be less than 1.
New matrix operation
nm number / NM number: a new matrix indexed as number is loaded, ranging from 1 to 6, which is the matrix QM number as in those buttons.

This java program is first created by Malte von Ruden, revised by Jie Wei and Ze-Nian Li.