Abstract: A method and apparatus to non-destructively inspect and quantify yarn crimp characteristics either off-line or on-line at speeds in excess of 2,000 yds./min. by acquiring images of yarn crimp on-line during the manufacturing process, then digitizing the images of yarn crimp and performing all data processing from the digital data required. Each image is segregated into features (e.g. valid crimp, low frequency oscillations, high frequency oscillations, edge oscillations, and unconfirmed oscillations with each parameter being analyzed separately. Features are segmented two dimensionally using techniques such as, but not limited to, local max/min intensity profiles and spatial intensity and frequency thresholding. Regions identified as abnormal are separated from areas of valid crimp before determining crimp frequency. Outputs of individual analysis are combined to determine global tow quality. This method provides a complete statistical analysis for each lane in the tow as well as the overall tow width.

Abstract: In an automated analyzer that measures concentration of samples by comparing changes in a detector response caused by said samples to a calibration curve or factor based on one or more different calibration standards stored in the analyzer, a replacement calibration standard is validated automatically by processing a portion of said replacement calibration standard as a sample prior to accepting the new calibration standard as a valid standard.