HK1020612B - Apparatus and method for characterizing fiber crimps - Google Patents

Description

Device and method for identifying fiber crimpness

Background

1. Technical Field

The present invention relates generally to an apparatus and method for measuring the crimp characteristics of fibers; more particularly, the present invention relates to an apparatus and method for characterizing fibers in a moving tow.

2. Introduction to the prior art

Manufactured or synthetic filaments are typically crimped into a bundle before being cut into staple fibers for further processing for various uses such as wool, sliver, or yarn. The fiber bundle is typically crimped by passing it through a crimping device to create a wave or crimp shape. Crimp characteristics such as crimp uniformity, number of crimps per inch of length, crimp frequency, etc. are often used to measure the quality of the manufactured fibers. Until recently, crimp characteristics were measured by manual inspection of cut portions of the fiber, such as counting the number of crimps per unit length.

Thus, an automatic crimp property measurement system would greatly increase the speed and accuracy of fiber identification, enable on-line adjustment of the production process, and produce staple fibers as desired.

Various automatic curl characteristic measuring systems have been proposed. These systems generally include a light source for illuminating a crimped fiber bundle; a camera for capturing an image of a portion of the crimped fiber bundle; circuitry for processing the captured image; and a display for displaying the measured curl characteristics.

One type of system employs a conventional TV camera to capture images of a moving, crimped fiber bundle. It will be readily understood by those skilled in the art that a TV camera takes images and converts these images to an amount of charge corresponding to the brightness of the moving crimped fiber bundle. These charge amounts are converted into video signals in order of pixels. The pixels are displayed on a monitor in an interlaced raster scan, e.g., the pixels are swept horizontally from top to bottom. In the interlaced scanning process, two fields are used.

After the first half frame scan from top to bottom is completed, blanking occurs when the scanning beam returns to the top, from which the scanning process is repeated for the second half frame scan. The second field scan lines are between the first field scan lines due to the deviation of the scan beam back to the beginning of the top of the raster by half a line from the beginning of the second field. Thus, the two sets of scan lines are interleaved. The two interlaced fields constitute a complete video frame.

There is a problem in acquiring images of a moving coiled beam using a common TV camera. The acquired image does not really represent the actual image because the fiber bundle is moving. Two fields interlaced with each other are obtained from different areas of the fiber bundle. When the scanning beam returns to the top of the raster to start the second field scan, the fiber bundle has moved and a different portion of the fiber bundle is scanned. Therefore, the measurement results are derived from the interlaced image having a deviation from the real image.

One popular way to solve the above-mentioned problems arising from the use of ordinary TV cameras is to employ a synchronized strobe system. A stroboscopic measurement light source emits light pulses that produce an image of the apparent stopped motion of the moving crimped fiber bundle. The cameras take snapshots of the moving curled beam synchronously as the light pulses are emitted. The synchronized strobe system effectively freezes the moving warp beam and the two interlaced fields no longer produce the erroneous image produced when using the non-strobe system. However, implementing synchronous control requires the use of advanced electronics. Furthermore, in order to cover the entire width of the moving fiber bundle when using a synchronized strobe system, it is necessary to move the focusing strobe light source and/or the camera, and to use a positioning mechanism such as a stepping motor and corresponding controls. If lighting or moving the camera is required, more time is required to position the devices, so it is not possible or difficult to achieve in-line or real-time measurement and/or device adjustment. In US patents US-4737846; 4415926, respectively; 4232336, respectively; and 4240110, some examples of strobe devices for measuring the crimp characteristics of fibers in a moving fiber bundle.

There is therefore a need for an apparatus for acquiring images of a moving crimped fiber bundle using a continuous or non-stroboscopic system that addresses the above-mentioned problems and enables on-line or real-time measurement and/or system adjustment.

Summary of The Invention

The present invention relates to an apparatus and method for measuring the crimp characteristics of fibers in a moving crimped tow in which a video camera is used to acquire video images of the moving crimped tow. The crimped fiber bundle is illuminated by a continuous light source as the camera acquires the image. A processor and associated software are used to decompose the acquired interlaced image into two non-interlaced images.

The processor and stored software converts the two non-interlaced images into a set of horizontal stripes. The set of strips is analyzed to measure the crimp characteristics of the moving crimped bundle portion represented by the strips. The measurements can be displayed and peripheral equipment adjusted to control deviations of the production process corrections from the operator specified specifications. The processor and stored program process the resolved images according to user-defined categories and display the frequency of crimps belonging to each category on a monitor to allow an operator to determine whether the crimped tow meets predetermined specifications.

The method of the present invention includes acquiring an image of a crimped fiber bundle; digitizing the acquired image; decomposing the digitized image into two field images; and a step of processing the two field images.

An exemplary method employed in the step of processing the two field images according to the present invention comprises the steps of: dividing one of the two images into a set of horizontal strips; establishing an intensity profile by averaging the pixel intensities for each band; identifying local minima and maxima in the intensity profile, and flagging a maximum as a curl peak if the difference in intensity between the maximum and its two immediately adjacent minima exceeds an operator-specified intensity threshold; calculating and storing distances between adjacent crimp peaks of all the peak values identified in the above step; these curl peaks are classified into a curl category.

A preferred method further comprises the step of transmitting the measurements to a set of peripheral devices configured for crimped fiber bundle production.

Preferably, the apparatus and method of the present invention also provide a convenient and reliable means of monitoring the quality of the crimped fiber bundle, and the physical data obtained can be used as a measure of quality control in the production process. For batch type production processes, the system also supports a start-up mode to minimize production rejection rates.

Preferably, an illumination means is also provided above said moving crimped tow for illuminating a portion of said moving crimped tow.

Brief description of the drawings

A preferred embodiment of the invention is described below with reference to the accompanying drawings, in which:

fig. 1 is a block diagram schematically illustrating a main part for measuring a fiber crimp characteristic according to a preferred embodiment of the present invention.

FIG. 2 is a flow chart illustrating one method for measuring fiber crimp characteristics in accordance with the present invention.

Fig. 3 is a flowchart of the start mode routine.

FIG. 4 is a flow chart of a "verification illumination" method.

FIG. 5 is a flow chart of a "Normal curl measurement" method.

Fig. 6a and 6b are monitor displays of curl criteria and system settings for a preferred curl measurement device constructed in accordance with the present invention in manual and automatic modes, respectively.

Fig. 7a, 7b, 7c, 7d and 7e are some exemplary setup displays for the automatic operation mode.

Fig. 8 and 9 are screen displays showing exemplary curl measurements in text and graphs.

FIG. 10 is a display of an exemplary alarm event.

11a, 11b, 11c and 11d are screen displays of analog and digital I/O test diagnostics for an exemplary curl measurement system constructed in accordance with the present invention.

Detailed description of the preferred embodiments of the invention

An exemplary crimp measurement system of the present invention is a digital image analysis system for identifying fiber crimp, including counting the number of crimps per unit length and crimp level distribution of a moving crimped fiber bundle. Such non-contact and non-destructive measurements may be taken off-line or on-line during yarn production, in which the crimped fiber bundle is moved at high speed, e.g., 1200 feet per minute.

Referring to fig. 1, the exemplary curl measurement system constructed in accordance with the present invention includes a computer 10, a frame grabber 12, an analog video monitor 14, a video signal switch board 16, a set of cameras 18, an I/O interface 20, and peripheral devices 22. The peripheral devices 22 include, but are not limited to, a crimping controller 24, a fiber bundle tension sensor 26, a light intensity adjuster 28, an external data memory 30, and an audio/visual alarm device 32.

The video signal switch board 16 may interface with the computer 10 and the set of cameras 18 via an RS-232 cable 34 to transmit video signals output from the set of cameras 18 mounted at different locations along the direction of travel of the fiber bundle or at different fiber bundle draw lines to the computer 10. The I/O interface 20 interfaces with a peripheral device 22 to enable I/O communications between the computer 10 and the peripheral device 22. Frame grabber 12 is coupled to an analog video monitor 14 and video signal switch board 16 to digitize the analog signals and transmit the digital data to computer 10.

Computer 10 is preferably an IBM compatible PC and includes a Pentium microprocessor and a Microsoft Windows operating environment. The computer 10 includes software in the form of a stored program 11 for controlling the hardware parts of the system. The VIDEO signal switch board 16 is preferably a KNOX VIDEO RS 12 x 2 type switch board capable of connecting up to 12 cameras. The operator uses the computer 10 to transmit signals through the RS-232 cable 34 to the video switch panel 16 to selectively receive video signals from one of the cameras 18. The selection method may employ a multiplexing technique commonly used by those skilled in the art. The I/O interface 20 is formed by two data acquisition boards, preferably a cyberrearch DAS 1601 board and a CYRDDA 06 board, for I/O communication between the computer 10 and the peripheral devices 22. The frame grabber is preferably a TARGA +64 digitizer capable of digitizing the video image into a two-dimensional data array.

The camera 18 is preferably a Panasonic GP-MF502 camera with electronic shutter speed control. The image acquired by the camera 18 is similar to that acquired by a conventional camera and is an interlaced image, i.e. a composite image of two (e.g. even and odd) field images. As previously described, each field image may represent a different area of the crimped tow due to the movement of the tow.

The user interface includes a main control panel, the measurements displayed on the computer screen, a keyboard or a mouse, and an icon on the computer display to enable the user to configure the curl measurement system, as discussed further below.

A light source (not shown) is provided adjacent to the set of cameras 18 for continuously illuminating the moving crimped fiber bundle. Preferably, the light source is a halogen flood lamp of sufficient intensity to cover the full width of the fiber bundle production line. The intensity of the light source may be adjusted by a light intensity adjuster 28, and the light intensity adjuster 28 is in turn controlled by the processor 10 and stored program 11. At least one tow draw line (not shown) is located below the light source to move the crimped tow. The exemplary section shown in fig. 1 may identify three or more tow lines. The tow draw line may be stationary, but it is preferred to move the crimped tow at a speed of about 1200 feet per minute.

Figure 2 shows the algorithm of the stored program 11 for measuring the crimp characteristics of a moving crimped tow. Under software control, the light source continuously illuminates the moving fiber bundle. The cameras 18 are preferably stationary and are capable of continuously capturing images of stationary or moving crimped fiber bundles, with signals representing the selected images being received by the computer 10 from one of the cameras 18 by selection of the video switch panel 16 under control of the stored program 11.

The selected image is digitized by frame grabber 12 and may be stored in computer 10 at step 100. In step 105, the received image is decomposed into originally interlaced even field and odd field images. The decomposition procedure is preferably performed by an image decomposition module in the stored program 11, as exemplarily shown in table 1.

Input of the Field _ Decompose function-memory pointer to the input image, height 400, width 512 output [2 ]]Two stored pointers for the output picture, field-0-field ID number, 0 for even field, 1 for odd field Loop: for RowIndex 1 to 400 Step 1 memory (output [ field ]]Input512) -copy image raw data from input image to output image output field]＝output[field]+512) -advancing the stored pointer for the output picture to the next line input + 512-advancing the stored pointer for the input picture to the next line If field 0, Then field 1 Else<dp n="d5"/>　　  field＝0　　End If　　Next Row Index

TABLE 1

The field decomposition function defines a loop to process each of the 400 lines of digitized data that make up a video frame. The processed lines are divided into a first field and a second field or an odd field and an even field according to the line index count. For example, when the field identification number in line 4 of Table 1 is set to 0, line 400 is separated from the video frame and designated as part of the even field, and the steps in the loop are repeated until all 400 lines are broken down into odd or even fields.

The decomposed image is noise reduced at step 110 using conventional image noise reduction techniques, such as using a filter. This process is optional and the disable can be set by the operator via the settings display of fig. 6a and 6 b.

At decomposition, there are 200 lines per odd and even field. To measure the curl characteristics, the rows in the odd and even fields are divided into M stripes, each stripe containing N rows. In this example, M is set to 50 and N is set to 4. Thus, each decomposed image has 200 lines, 50 stripes, each of which contains 4 lines.

Each band is averaged at step 115 to form an intensity profile, which is represented by a gray scale from 0 (black) to 255 (white). All local maximum and minimum curl peaks in the intensity profile are located according to the operator specified curl intensity threshold at step 120. If the intensity of one crimp peak differs from its two adjacent crimp peaks by more than an operator-specified intensity threshold, then that crimp peak is marked as a maximum. The curl strength threshold is adjusted based on optical factors associated with the fabric material that take into account the amount of light absorption and light reflection by the fabric material.

The display input curl strength threshold is set by curl measurement as shown in fig. 6 a. The crimp strength threshold in fig. 6a has been set to 8.

The frequency of adjacent maximum curl peaks is calculated in step 125, where the frequency is defined as the inverse of the distance between two adjacent maxima. At step 130, a maximum is flagged as a valid curl if its corresponding frequency is within the user specified range.

Each curl is classified into one of three predetermined categories, which include micro-curl, normal curl, and macro-curl. The classifications are determined by one of the system measurement setting displays shown in fig. 6a and 6b according to the CPI ranges specified by the operator for each of the three classifications. For example, in fig. 6a and 6b, if the CPI parameter for a curl is greater than or equal to 16, 8, or 4, respectively, it is classified as a micro-curl, a normal curl, or a macro-curl.

The measurements were statistically analyzed to determine parameters such as average CPI and the percentage of area covered by each type of curl. The statistical analysis is displayed on a computer screen to enable the operator to see the data, as described in detail below.

Steps 115 through 130 are repeated for each strip until all image lines in a decomposed image have been analyzed. Steps 115 to 130 are then performed to analyze the curl characteristics of the second non-interlaced image (140).

More than one camera may be used in order to capture an image representing the entire width of the crimped fiber bundle, which is typically about 4 inches or more. In this embodiment, three cameras are used for each tow line. The same steps 100 to 140 shown in fig. 2 are performed in analyzing the images acquired by the other cameras to obtain the crimp characteristics of the entire width of the moving crimped fiber bundle. The data from all three cameras were averaged to obtain the overall curl characteristics results. These results are displayed on a computer screen along with other statistical analysis results to enable an operator to check the crimp characteristics of the crimped fiber bundle.

Fig. 8 and 9 are examples of curl measurement result screen displays. Fig. 8 shows that the average CPI for all fiber bundles was 9.8 and that for the micro-, normal-, and macro-crimp types, the crimp CPI was 22.6, 11.2, and 6.2, respectively. Figure 8 also shows that the measurable crimp covers 69.8% of the area of the fiber bundle and 3.0%, 40.5%, and 26.3% of the area covered by crimps of the micro-, normal-, and macro-crimp species, respectively. The right side of the screen display of fig. 8 also includes a main user interface control panel for controlling the curl measurement system and setting system parameters.

Fig. 9 shows the statistics of the in-line crimp distribution for a crimped fiber bundle. These distribution statistics are constantly changing as the curl measurement program progresses to enable an operator to continuously monitor the curl measurements of the moving crimped fiber bundle. This online curl distribution screen display also indicates the moving average of curls (mAvg), the percentage of each curl category covering area and the total area covered by the curl, and other data.

An operator viewing these measurements can instruct the peripheral device 22 via the I/O interface 20 to take appropriate action to enable the crimped tow production process to meet the appropriate product and process requirements. As one example, the operator may instruct the crimper controller 24 to increase or decrease the amount of crimp or reconfigure the production process to re-determine the number of crimps divided into micro crimps, normal crimps and large crimps depending on the purpose of crimping the tow.

The operator may select a manual or automatic mode of operation, as shown in the upper left side of fig. 6a and 6b, with reference to the system measurement setup display shown in fig. 6a and 6 b. In fig. 6a the manual operation mode is selected, whereas in fig. 6b the automatic operation mode is selected. Other initial settings or adjustments to the curl measurement system may be accomplished with the help of the settings displays shown in fig. 6a-6 b. For example, the image resolution, stripe size, curl strength threshold, effective Curl Per Inch (CPI) range, overall CPI set point, CPI tolerance; selecting a fiber bundle drawing production line and a camera; crimp type and specifications; and the image preprocessing configuration performs parameter adjustment or setting.

At the lower left side of fig. 6a and 6b, the operator can choose whether to smooth the desired image before processing. If smoothing is selected, each of the odd field image and the even field image is passed through a filter to subject the images to noise reduction processing as described above. The number of scanning lines for an image constituting a strip can also be selected in the same dialog box labeled "image preprocessing". The selected stripe size is 8 in fig. 6a and 4 in fig. 6 b.

In the system measurement setting display shown in fig. 6b, in which the automatic operation mode has been selected, the number of fiber bundle draw lines selected to perform the measurement is 3. The number of cameras selected for each tow line is also 3. Below the selection of the number of fibre bundle draw lines and the number of cameras for each fibre bundle draw line is a user interface for entering a system setup display of the automatic operation mode shown in figures 7a-7 e.

The General setup dialog of the automatic mode setup display shown in FIG. 7a may be entered by selecting the box labeled "General". This setup display enables the operator to set the overall system setup and various parameters such as sampling rate, number of images held on the screen, number of moving average data points, image resolution, fiber bundle tension adjustment factor, crimp strength threshold, fiber adjustment factor, average image strength, tolerance factor, and no-fiber bundle image strength.

The generic name setting display shown in FIG. 7b may be entered by selecting the block labeled "Alias" in FIG. 6 b. This setup display enables the operator to provide one short name and one long name for each fiber bundle draw line and for each camera across its width located on these fiber bundle draw lines. As shown in fig. 7b, the short and long names of the fiber bundle drawing line are 0, 1 and 2, and ts800, ts801 and ts802, respectively. The short and long common names of the three cameras are R, C and L, respectively, and the right, center and left cameras.

The Trend window setup display shown in fig. 7c may be entered by selecting the box labeled "Trend" in fig. 6 b. This setup display allows the operator to select the parameters he needs to display during normal measurements. The selected parameters are displayed on an on-line curl trend window, as shown in FIG. 9. Using the trend window settings display, the operator can set the range of the overall CPI parameters, the percentage of the area covered by the overall CPI, and the percentage of the area covered by the micro-, normal-, and macro-curl-like curls.

The I/O usage settings display shown in FIG. 7d may be entered by selecting the box labeled "I/O" in FIG. 6 b. The display of fig. 7d allows the operator to manually activate each draw line, start-up sequence, bad tow alarm, specification alarm, lighting, tow tension enable and disable, and set the overall CPI range and other parameters. On the right side of fig. 7d, the operator can activate the system fault alarm and perform digital or analog diagnostic tests on each data acquisition board.

The Start setting display of FIG. 7e may be entered by selecting the box labeled "Start Up" in FIG. 6 b. This display allows the operator to configure the start mode program, which is another advantageous feature of the device, as described below. From fig. 7e, the operator can set the image resolution, ribbon size, curl strength threshold, minimum measurable area, effective curl range, and other start-up mode program parameters.

The start mode procedure is represented by fig. 3. This start-up mode procedure is particularly advantageous when used in a batch-type fibre production process, i.e. the staple fibres are supplied in batches and are usually broken after a certain length. The fiber characteristics of the beginning portion of each batch of fibers are typically "featureless" or different from the rest of the batch. This is mainly due to loose ends and other imperfections formed by the cutting. These featureless beginning portions are typically cut away and discarded.

The start mode is used to monitor the initial portion of each batch and alert the operator when the crimped tow reaches the "characterized" portion of the batch, whereupon the operator cuts and discards the initial "featureless" portion. The startup procedure reduces waste due to unnecessarily discarding too long a starting portion and allows each batch to have more consistent characteristics. The start-up mode also prevents possible corruption of data by preventing mixing of the measurement of the crimp characteristic of the start portion of the crimped tow with the measurement of the main portion of the tow. The start mode triggers a signal when the good tow is in the direct field of view of at least one camera 18 used to image the entire width of the moving crimped tow.

Referring to fig. 3, the curl measurement system is activated by turning on the continuous light source (300 a); setting a draw line flag and turning on a bad tow signal light (300b) to indicate the start of a batch; selecting a startup video channel (300c), the camera being a stand-alone camera specifically configured to capture startup images; setting a fiber bundle and an interruption time counter (300 d); and prepares a trend window (300 e).

The interrupt time counter is used to set the time to complete the start-up procedure. The counter is updated (305) and checked to determine if it has reached a predetermined interrupt time (310). When the interrupt time is reached, the startup mode is terminated. The illumination is turned off (315a) and the bad tow signal lamp is turned off (315b), and normal measurements are resumed (315 c).

During the interruption time, an image of the moving crimped fiber bundle is acquired using the start-up camera (320). The acquired image is displayed on a screen (325). Curl Per Inch (CPI) parameters are measured and recorded and a trend window is updated (335).

The illumination is checked by computer 10 to determine if the average image intensity is within specification (340). If the illumination does not meet the operator specification, the image intensity is increased or decreased by increasing or decreasing the voltage of the continuous light source (405). If the light source voltage exceeds a specified level (410), a system alarm (415) is checked and an alarm signal is triggered in the following steps. An alarm/event message (420) is displayed on the computer screen to indicate lighting problems or other system problems to the operator. Fig. 13 is a screen display with an alarm/event information window.

If it is determined at step (425) that the system alarm I/O has been validated, a system alarm lamp (430) is turned on. System alarm I/O enablement is accomplished via the system settings display shown in FIG. 7 d.

If the lighting does not meet the operator requirements, the process returns to step 345 of FIG. 3. The software checks to determine if the measured crimped tow region is within operator demand at step 345. If not, the routine returns to step 305 and updates the interrupt time counter. If the software determines at step 345 that the measured area of the crimped tow is within the operator's requirements, then the featureless starting portion of the tow may have moved through the system.

At step 350, the stored program 11 checks to determine if the CPI parameter of the crimped tow is within the operator desired range. If the CPI parameter is within the operator request, the good tow counter is updated (355). The good tow counter is checked to determine if a good tow count has been reached (360). If the good tow counter has reached the good tow count, the routine returns to step 315 to terminate the Start mode function. The start mode function is terminated by turning off the illumination (315 a); turning off a bad tow warning light to indicate to an operator that a good tow has been reached (315 b); and normal curl characteristic measurement is resumed (315 c).

If it is determined at step 350 that the CPI parameter is not within the operator desired range, the tow counter is reset (365) and the start-up routine returns to step 305 to update the interrupt time counter. Likewise, if a good tow count has not been reached, the process returns to step 305. The startup sequence continues until the CPI parameter is determined to be within operator requirements at step 350 and a good tow count has been reached at step 360.

The flow chart of fig. 5 shows the interaction between software and hardware components and status checking during normal measurement mode. Steps 500 and 505 determine whether a start trigger has been triggered to switch to the start-up sequence mode shown in the flow chart of fig. 3. If the start trigger has been triggered, for example in the case of batch-type processing, the normal measurement is postponed and the start-up procedure is continued.

In the normal measurement mode, computer 10 and associated program 11 determines whether camera 18 is capable of capturing images of the moving crimped fiber bundle. Preferably, the system uses a camera positioned to move the crimped fiber bundle across its width to image the entire width. If one of the three cameras is disabled or has failed to function properly, the measurement continues with the other camera. According to a preferred embodiment of the invention, three different tow draw lines may be run simultaneously using three cameras across the width of each tow.

The number of fibre bundle draw lines measured in the automatic mode of operation is set by means of the setup display shown in figure 6 b. The storage program 11 checks the camera 18 and the cutter before taking the selected image. For example, when one camera 18 is enabled (515), the slicer I/O is not enabled or the slicer is operating (steps 520 and 525), images are acquired for processing with the selected camera (step 530).

It should be noted that although three cameras are selected in the present embodiment, the curl characteristic measurement may be performed using images acquired by fewer than three cameras. In some cases, the operator may not enable one of the cameras. For curl measurement and illumination control, the average intensity of all the images acquired by the active cameras is used.

If the camera is not enabled in step 515, then a next video channel is selected in step 535 and the processor 10 and stored program 11 receive images of the moving crimped fiber bundle acquired by the selected video channel camera (step 540). If no image is received, the process returns to step 515.

The captured image is displayed in an image window (545); measuring, displaying and recording a CPI parameter (550); and a moving average is calculated (555). The measurement may be compared (560) to the operator request and the specification status checked (565), the information window updated (570), and a flag may be set to indicate that the specification status has changed (575).

If the specification status has not changed or after the flag has been set indicating that the specification status has changed (580), the stored program 11 checks whether all enabled cameras on a tow line have been selected for sending images of moving crimped tows. If not, steps 515 through 580 are repeated for another camera. After all enabled cameras have transmitted images, the program checks the illumination based on the average image intensity of the images acquired from the three cameras and adjusts the illumination control if necessary (steps 585, 590, and 595. see also fig. 4).

Curl measurements are also based on averaging the images acquired by the three cameras. For example, if CPI output I/O is enabled (step 605), the average CPI is output and may be displayed (step 605). The storage program 11 checks at step 610 whether the measured curl characteristic indicator has been different from previous measurements and updates the trend window accordingly at step 615 so that the operator can see the measured on-line curl characteristics and other measured parameters. An alarm may be used to alert the operator whether the measurement is outside of a predetermined specification.

The procedure shown in fig. 5 is repeated for additional tow lines.

Preferably, the use of a continuous wavelength light source to illuminate the moving crimped fiber bundle and a conventional TV camera can eliminate the need for a stepper motor to control the moving light source or camera. The curl characteristic measurement (of the present invention) can be displayed at near real-time speed compared to systems using stepper motor control. In previous systems, the updating of the measurement results was very slow due to the time required to position the lights and cameras. Accordingly, system parameter adjustments are correspondingly delayed in systems that use motor positioning control.

The stored program 11 also provides a method for examining the data acquisition board through a diagnostic test as shown in the screen displays of fig. 11a-11 d. These screen displays allow the operator to set the position/channel location, for example. Channel logo, and control inputs and outputs.

It will be understood that various modifications may be made to the embodiments of the invention without departing from the spirit thereof. The above description is not intended to limit the invention but merely to illustrate preferred embodiments thereof. Those skilled in the art will envision other modifications within the scope and spirit of the invention as defined by the claims appended hereto.

Claims (10)

1. A system for measuring the crimp characteristics of fibers in a moving crimped tow, said system comprising:

a processor and associated stored program;

an illumination means positioned above said moving crimped tow for illuminating a portion of said moving crimped tow;

at least one camera for capturing at least one interlaced video image of said portion of said moving crimped fiber bundle;

digitizing means for digitizing said at least one interlaced video image into digital data;

decomposition means for decomposing said digital data into decomposed data representing first and second non-interlaced field images, wherein said processor and associated stored program processes said decomposed data; and

a display for displaying curl characteristics based on results of said processing of said decomposed data.

2. The system of claim 1 wherein said decomposing means includes means for dividing said first and second field images into a plurality of horizontal bands and for creating an intensity profile for each of said bands by averaging pixel intensities of successive horizontal scan lines within each of said bands.

3. The system of claim 1, wherein said processor processes said resolved data into minimum and maximum intensity profiles, wherein a maximum is labeled as a curl peak if the difference in intensity between the maximum and its two adjacent minima exceeds an operator specified intensity threshold.

4. The system of claim 1 wherein said processor calculates a distance between adjacent curl peaks, compares said distance to an operator specified threshold, classifies said curl peaks as one of a micro-curl, normal curl, or macro-curl category, and tabulates overall curl characteristics of said first half-frame image.

5. The system of claim 1, wherein said processor communicates said measurement to at least one peripheral device to configure said system according to predetermined specifications.

6. The system of claim 1 including a plurality of cameras covering the full width of said moving crimped fiber bundle.

7. The system of claim 6 wherein said processor and stored program control a video switch board to selectively receive signals from one of said plurality of cameras.

8. The system of claim 1, wherein said illumination device is a continuous wavelength light source.

9. The system of claim 8 wherein said illumination means illuminates at least a full width of said moving crimped fiber bundle.

10. A method for measuring the crimp characteristics of fibers in a moving crimped tow, said method comprising the steps of:

a) illuminating said crimped fiber bundle with a continuous wavelength light source;

b) capturing interlaced video images of the crimped fiber bundle;

c) digitizing the video image;

d) decomposing the digitized image into two non-interlaced field images;

e) processing said two field images; and

f) curl characteristics are exhibited based on the results of said processing of the image.