Small container fluid dynamics to produce optimized inspection conditions

• US 7,430,047 B2
• Filed: 03/09/2005
• Issued: 09/30/2008
• Est. Priority Date: 03/09/2004
• Status: Active Grant

• Current Assignee: Gerald Walter Budd, Julius Z. Knapp
• Original Assignee: Gerald Walter Budd, Julius Z. Knapp
• Inventors: Knapp, Julius Z., Budd, Gerald Walter
• Primary Examiner(s): Punnoose; Roy M
• Application Number:
  US11/076,375
• Publication Number:
  US 20050248765A1
• Time in Patent Office: 1,301 Days
• Field of Search: 356/427, 356/335, 356/338
• US Class Current: 356/427
• CPC Class Codes: G01N 15/1459 the analysis being performed on a sample stream
• CPC Class Codes: G01N 2015/1027 Determining speed or velocity of a particle
• CPC Class Codes: G01N 2015/1493 Particle size
• CPC Class Codes: G01N 21/9027 in containers after filling

Abstract

New methodology, realizable with both manual and new semi-automatic imaging technology, has transformed both the inspection and the batch release Attribute Sampling Inspection for contaminating visible particles in injectable solutions into statistically replicable procedures. In this new non-destructive inspection procedure, a calibration curve relates NIST traceable measurement of maximum particle size to the rejection probability of the particle. Data for this calibration curve is determined with a graduated set of single durable stainless steel and glass microspheres that are sized with NIST traceability. Use of the calibration curve transforms the probabilistic variability of visible particle inspection data described by Knapp into the ‘simply replicable form’ required by the Attribute Sampling Tables. The present invention uses cutting edge imaging technology to achieve 1% sizing accuracy within 10 μm from 50 to 1,000 μm. An improved alternative sizing technique used in this invention uses the particle information to achieve an integral particle sizing function.

Claims

1. An improved apparatus with diffuse backlight illumination system for the accurate detection and measurement of key characteristics of moving particles in the solution of small containers, 1 ml to 50 ml, with a machine vision measuring system comprising:

a) an image processing computer for image acquisition, image storage, image processing capability and analysis of images or derivations of images;

b) the image processing computer comprising memory for storing the images formed by the camera or cameras;

c) the image processing computer also comprising digital parallel input/output digital serial, and Ethernet communication capabilities for providing messages to external devices to report one or more measurements or characteristics of the particles in the solution that have changed in position the stored images;

d) the image processing computer executing control software stored in a computer readable medium, for allowing request and response signals from external devices indicating the parameters required for the inspection of a specific container type, for causing the image processing computer to perform image alignment and analysis for extraction of key characteristics of particles in the solution that have changed position within the container;

e) the image processing computer executing control software stored in a computer readable medium, for allowing request and response signals from external devices causing the image processing computer to store a reference image of container and the solution within the container without the detection of any contaminating particles in the solution in a memory location referenced by a specific identification code that is unique to the container model, type, specific contents and fluid fill level;

f) the image processing computer executing control software stored in a computer readable medium, for allowing request and response signals from external devices causing a determination of the exact position of the container by extracting one or more edges of the container based on grayscale information for causing the image processing computer to accurately position the inspection zones;

g) one or more image sensors with lens that provides a spatial resolution of 15 to 25 micrometers per pixel and depth of field necessary to form a sharp focus image of substantially all of the bottom interior surface of the container or a target portion thereof including the meniscus;

h) wherein each image sensor comprises the pixel resolution to resolve a contaminating particle of at least 20 micrometers in diameter resting on the interior bottom center of the container;

i) a minimum of one set of optical components are aligned so that the optical path is at a downward angle less than perpendicular from the axis of rotation permitting the sensor to view substantially all of the bottom interior surface of the small container;

j) an optional second set of optical components are aligned so that the optical path is at an angle perpendicular from the axis of rotation permitting the sensor to view half of the fluid volume and half of the lower surface of the fluid meniscus within the container;

k) an illumination system comprised of a cube structure with a “

U”

shaped channel cut into an optical grade polycarbonate or acrylic cube, and is machined so that the center of the “

U”

channel coincides with the axis of rotation of the container, with the diameter of the channel allowing the container being inspected to fit inside without interference with the channel walls;

l) the illumination system implements high density of light emitting diodes (LED'"'"'s), a minimum of 4 LED'"'"'s per square centimeter, arranged around the perimeter of the diffusing cube and is positioned to uniformly illuminate the contents of the container;

m) the illumination system utilizing one or more precision power supplies with control circuitry to turn on or off sections the LED'"'"'s as required by the image processing system to enhance the contrast of the particles in the solution of the container;

n) a precision drive motor is connected directly to recessed bottom container holder is used to impart rotational motion to the base of the container in a precise and repeatable manner with a motion processing computer and drive unit comprising memory for storing one or more defined motion programs;

o) the motion processing computer also comprising digital parallel input/output digital serial, and Ethernet communication capabilities for providing messages to external devices to report status of operation;

p) the rotational motion limited to insure the angular acceleration and velocity do not cause deformation of the meniscus from a smooth parabolic shape;

q) the rotational motion of the container imparts a toroidal motion to the fluid in the container that will pull contaminating particles in the fluid toward the axis of rotation;

r) whereby image sensors are positioned relative to the axis of rotation of the container, whereby a focal point of detection coincides with the axis of rotation of the container so to view the content of the solution in the container and specifically the interior bottom center of the container with the illumination system providing a contrasting geometric size and shape of substantially all of the particles in the solution that have changed position in the image being identified, binary and grayscale information being recorded, particles evaluated and measured, and reporting the information of each particle on being displayed on a human machine interface.

4. The method of claim 1, wherein the time required for the inspection, including image acquisition, image storage, image processing and analysis for each of the container'"'"'s one or more defined inspection volumes is between 5 and 30 scan times;

wherein a typical scan time is 30 to 60 milliseconds.

5. The method of claim 1, wherein the detection of contaminating particles is based on either a) light extinction that produces well-defined geometric shape(s) or b) grayscale shapes defined by the summation of grayscale values in an area defined by the approximate geometric shape.

6. The method of claim 5, wherein the aspect ratio of the grayscale shape is in a range of 0.8:

1 to 1.2;

1 and defined by geometric width and an equivalent number of pixels in the vertical dimension starting from the geometric top of the shape.

7. The method of claim 5, wherein the measurement of particles is confined to the defined inspection volumes, and is further confined to the front two-thirds of the container volume nearest to the detector, and is further confined to eliminating the extreme zones toward the far left and far right near the container walls that are more than 75 percent of the distance from the centerline of the container to the container walls.

8. The method of claim 7, wherein the measurement of small particles (low to medium mass, equivalent to <

=350 μ

m diameter stainless steel spheres) is further confined to those particles in a defined inspection volume centered about the bottom center of the container, that is no more than 50 percent of the distance from the centerline of the container.

9. The method of claim 7, wherein the measurement of large (heavy mass, equivalent to >

350 μ

m diameter stainless steel spheres) particles is further confined to those particles that travel from left to right in the images, assuming a counter-clockwise rotation of the container, in a defined inspection volume near the forward portion of the container and more than 75 percent of the distance from the centerline of the container.

10. The method of claim 7, wherein the measurement of light particles (low density <

=density of the fluid) is further confined to those particles that move in the solution through an inspection volume centered in the container but not including the bottom of the container.

2. An improved method for detection, measurement and determination of key characteristics of contaminating particles found in the solution within a container (1 ml to 50 ml) by placing substantially all particles within a predetermined size range, contained in the solution, into motion without turbulent flow using a precise velocity motion profile (VMP) designed specifically for causing the fluid contents of the container of a given size and shape, fluid surface tension, fluid viscosity and fill level to be set in motion, utilizing a programmable rotary motion device, an image processing computer for image acquisition, image storage and image processing capability for executing control software embodied on a computer readable medium, to provide the steps of:

a) use of a computer data file for extraction of key inspection features for a specific inspection parameters in a memory location referenced by a specific identification code that is unique to a container model or type that define the dimensions of the container, contents of the container, and the fill level of the container;

b) use of a computer data file for extraction of key inspection features for a specific inspection parameters in a memory location referenced by a specific identification code that is unique to a container model or type that define the dimensions of the container, contents of the container, and the fill level of the container;

c) storing one or more reference inspection mask images of a new master data file in a memory location referenced by a specific identification code that is unique to a container model or type that define the dimensions of the container, contents of the container, and the fill level of the container;

d) storing one or more rotary motion programs called velocity motion profiles (VMP) of a new master data file in a memory location referenced by a specific identification code that is unique to a container model or type that define the dimensions of the container, contents of the container, and the fill level of the container;

e) the velocity motion profile defines an acceleration, a maximum rotational velocity, a duration at maximum rotational velocity, followed by a deceleration to zero rotational velocity;

f) the velocity motion profile imposes a toroidal flow of the solution in the container; and

g) the fluid flow generates negligible or no turbulence in the meniscus;

h) characterized in that motion is imposed on substantially all particles in solution with a mass equal to or less than, the upper limit particle mass defined by the velocity motion profile given the attributes of the container type parameters;

i) characterized in that motion is imposed on substantially all particles in solution with a size equal to or less than the upper limit particle size defined by the velocity motion profile given the attributes of the container type parameters;

j) one or more velocity motion profiles are utilized for the inspection for a specific container type, fluid content and fluid fill level;

k) acquisition of a test image of a sample referenced by a specific identification code that is unique to a container model or type that defines the dimensions of the container, contents of the container, and the fill level of the container and storage of the test image in memory;

l) determining the size and position of one or more inspection zones of the grayscale image as referenced by a specific identification code that is unique to a container model or type that defines the dimensions of the container, contents of the container, and the fill level of the container and storage of the inspection zones in memory;

m) multiple inspection zone are implemented simultaneously;

n) placement of a test container in the inspection position, excitation of the container with a specific velocity motion profile, acquisition of a sequence of images and storage of the images in the image processing computer at specific time from the start of the velocity motion profile or alternatively from the end of a velocity motion profile as required by the specific identification code that is unique to a container model or type that defines the dimensions of the container, contents of the container, and the fill level of the container;

o) creating and storage of a reference background image for the first 50% of acquired images by using the last acquired image in the total sequence of acquired images;

p) creating and storage of a reference background image for the last 50% of acquired images by using the first acquired image in the total sequence of acquired images;

q) creating a difference image by grayscale subtraction of a test image from the reference background image as determined in steps o) and p);

r) creating and storing in memory a list of values comprising grayscale objects representing resulting information in the difference image that are larger than a pre-determined criteria;

s) reducing the number of grayscale objects in the list created in step r) by applying pre-determined size and shape filtering algorithms and then storing in memory the remaining objects in a revised list;

t) containers that have objects on the revised list are defective and the image processing computer generates an error message prompting a user that the container tested is defective and requesting the removal of the container from acceptable product;

u) the inspection information from each test is displayed on a human machine interface and stored permanently on one of the internal storage hard drives or written to a recordable digital video disc for offsite record storage.

13. The method of claim 2, wherein the utilization of a specific velocity motion profile (VMP) for a given container shape, size, fluid fill level, fluid viscosity imposes a toroidal flow causing separation of particles by mass, where the smaller particles are moved to the center of the container and the sensor has maximum focus to detect the smallest particles.

14. The method of claim 2, wherein the invention performs simultaneous independent evaluations of the near field and far field regions of the images and applies individual inspection criteria to each zone.

15. The method of claim 2, wherein the invention is used for counting the number or particles and determining the sizes of those particles in solution in the container, and creating a histogram of particle size verses number of particles in each bin on the human machine interface display.

16. The method of claim 15, wherein the invention is used to characterize the particle population in the 40 μ

m to 1,000 μ

m size range in a individual container or a set of containers.

18. The method of claim 2, wherein large contaminating particles with a density greater than that of the fluid, such as glass shards, metal chips and debris can be detected and measured between 2 and 10 scans in the inspection volume that includes the bottom of the container and occurs within 300 milliseconds of the end of the velocity motion profile (VMP).

3. An improved method for the creating an instrument calibration curve that allows the correlation of the dimension of a particle contained in the solution of a transparent container (1 ml to 50 ml) as determined by image acquisition and processing to that of NIST traceable particles standards, within a specified size range, comprising the steps of:

a) the movement of particles in the container whereby rotation of the container using a predefined velocity motion profile causing substantially all of the particles in the solution in the container to experience a change in position with a corresponding change in time;

b) recording the properties of particles isolated in each defined inspection volume;

c) determine the mean of the particle properties in each defined inspection volume over “

n”

images, primary properties included the geometric shape and integration of grayscale values over the geometric shape;

d) evaluate the particle only if its properties lie within a specific range of the mean of “

x”

samples and establish a mean equivalent pixel width in each corresponding inspection volume;

e) the generation of a calibration table by evaluating a series of NIST traceable dimensional standards in which the standards increase in diameter and recording the equivalent pixel width of the particle in each corresponding inspection volume;

f) the generation of a calibration curve by plotting the pixel width on the x-axis and the NIST standard diameter on the y-axis to produce a linear equation in the form of Y=mX+b;

g) one or more numerical calculations are performed to correlate particle size in pixels to equivalent physical dimension in micrometers using calibration curve established by measuring NIST traceable single seeded particle standards in each of the defined inspection volumes.

11. The method of claim 3, wherein a series individual NIST traceable spheres of known diameters, diameters range from 40 μ

m to 400 μ

m, each sample placed in a thoroughly cleaned container, filled with WFI (ultra pure, Water for Injection) to the proper level, and measured to determine equivalent pixel diameters to generate a calibration curve to be used in conjunction with the bottom center inspection volume.

12. The method of claim 11, wherein the calibration curve is used to determine the equivalent diameter in micrometers of contaminating particles in other containers of the same shape, size and fluid fill level when said invention determines diameter in pixels.

17. The method of claim 3, wherein the method of using the mean of multiple measurements provides an accuracy improvement in determination of a contaminating particle size which is displayed on the human machine interface.