Data compression system

• US 4,777,620 A
• Filed: 02/20/1985
• Issued: 10/11/1988
• Est. Priority Date: 02/20/1985
• Status: Expired due to Fees

• Current Assignee: Elscint Limited (Elbit Imaging Limited)
• Original Assignee: Elscint Limited (Elbit Imaging Limited)
• Inventors: Nissenson, Bilha, Shimoni, Yair
• Primary Examiner(s): Shaw, Gareth D.
• Assistant Examiner(s): Nguyen, Viet Q.
• Application Number:
  US06/703,434
• Time in Patent Office: 1,329 Days
• Field of Search: 364/514, 364/900, 364/200, 364/724, 333/14, 382/54, 382/56
• US Class Current: 382/233
• CPC Class Codes: H04N 19/63 using sub-band based transform, e.g. wavelets

Abstract

A data compression-expansion (decrease-increase) method for improving the signal-to-noise ratio of compressed-expanded data while retaining the fidelity of the original data comprises the steps of using a filter that can be separated into a smoothing portion and an enhancing portion. The selected filter is separated into the two portions: a first portion for smoothing prior to decreasing the data, and a second portion for enhancing after increasing the data whereby the filter has the maximal effect of decreasing noise while having a minimal effect on the data.

Claims

1. A data compression-expansion method for use in digital subtraction angiographic systems for improving the signal to noise ratio of the expanded data while maintaining the fidelity of the data;

said method comprising the steps of;

acquiring data,

digitizing the acquired data to provide uncompressed digital data,

using a transfer function separable digital filter which filters out noise and has a minimal effect on the digital data, said separable filter comprising a filter that has frequency response of approximately 1 over the applicable frequency range with a slight fall off at the high end of the range, said separable filter having a first portion and a second portion,

filtering the uncompressed digital data to smooth the uncompressed digital data with said first portion of the separable filter,

compressing the smooth digital data by coding to reduce the quantity of the smooth digital data while maintaining the amplitude of the retained data,

operating on the compressed smooth digital data,

expanding the compressed smooth digital data after said operating step to restore the quantity of the smooth digital data present prior to the compressing step, and

inherently restoring most of the digital data removed by the first portion of the separable filter by filtering the expanded digital data with said second portion of the separable filter to enhance the expanded digital data.

2. The data compression-expansion method of claim 1 wherein the step of operating on includes storing said smoothed, compressed data.

3. The data compression-expansion method of claim 1 wherein the step of operating on includes transmitting said smoothed, compressed data.

4. The data compression-expansion method of claim 1 wherein the step of operating on includes processing said smoothed, compressed data.

5. The data compression-expansion method of claim 1 wherein the data is two dimensional and wherein the step of filtering the uncompressed digital data comprises filtering in two dimensions.

6. The data compression-expansion method of claim 5 wherein the step of filtering the expanded data comprises filtering in two dimensions.

7. The data compression-expansion method of claim 1 wherein each portion of the separable filter includes a one-dimensional filter operating along the X axis and a one-dimensional filter operating along the Y axis.

8. The compression-expansion method of claim 1 wherein the first portion of the separable filter has a wavelength reducing factor of Gn, where n=2 and Gn is smaller than twice the expected amplitude of blood vessels while larger than the square root of twice the expected amplitude of the noise in the system whereby the ensuing reduction in noise offsets the noise addition of a subsequently used sharpening filter.