Browse Prior Art Database

Noise Insensitive Step Detection Method

IP.com Disclosure Number: IPCOM000078813D
Original Publication Date: 1973-Mar-01
Included in the Prior Art Database: 2005-Feb-26
Document File: 2 page(s) / 57K

Publishing Venue

IBM

Related People

Schechtman, BH: AUTHOR

Abstract

In the analysis of data streams generated either by numerical computation or, in the case of sensor based computer applications, by acquisition of data from physical equipment, there is frequently a need to scan the data stream for the purpose of locating a stepwise increment in the value of some variable of interest. The technique described here and illustrated in the figure presents an efficient method for achieving such step detection, even if the desired data Increment is masked by the occurrence of gradual drift and or random noise in the data stream, as frequently occurs in real-time laboratory automation or process-control applications. This discrimination against noise and drift is accomplished, by testing selected pairs of points in the vicinity of the initial detection of the step.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 63% of the total text.

Page 1 of 2

Noise Insensitive Step Detection Method

In the analysis of data streams generated either by numerical computation or, in the case of sensor based computer applications, by acquisition of data from physical equipment, there is frequently a need to scan the data stream for the purpose of locating a stepwise increment in the value of some variable of interest. The technique described here and illustrated in the figure presents an efficient method for achieving such step detection, even if the desired data Increment is masked by the occurrence of gradual drift and or random noise in the data stream, as frequently occurs in real-time laboratory automation or process-control applications. This discrimination against noise and drift is accomplished, by testing selected pairs of points in the vicinity of the initial detection of the step.

The present technique incorporates the above pairwise testing method into a high-reliability step detection scheme, which is useful in a variety of applications in which the criteria for determining the height of the step, its location and sharpness, and the tolerable background drift may all be variable. Once an increment in the data stream is found, which satisfies all of these criteria under a given set of specifications, the step location and magnitude are returned to the invoking program and normal processing of the data continues.

Referring to the figure, the sharpness of the step sought is defined through a bandwidth parameter (J...