MACHINE VISION BASED AUTOMATIC NEEDLE CANNULA INSPECTION SYSTEM AND METHOD OF USE

Abstract

A machine vision based automatic needle cannula inspection system includes an inspection and control unit, image capture devices, light sources, a unit that makes the needle cannula and the image capture device(s) rotate relative to each other, and a rejected part removal unit. By means of rotating the needle cannula and image capture devices relatively, a plurality of images captured along the circumferential direction of the needle cannula are directly saved to a computer, the images are then screened, processed and analyzed to fulfill the automatic inspection of multiple quality and technical parameters of the needle cannula without the need to position the bevel area of cannula tip to a specific direction. Inspection parameters and accuracy can be set at any time, the system can automatically record classification and statistics of passed and rejected needle cannulae for query, and the rejected cannulae are removed automatically at next position. All the functions are completed at one work station, and needle cannula tips on both sides can be inspected simultaneously, if needed. Thereby the inspection efficiency and reliability are greatly improved, resulting in extremely extensive application prospects and huge economic values.

Claims

1. A machine vision based automatic needle cannula inspection system comprising: a control unit configured to perform synchronized control and information processing; a needle cannula retainer for retaining at least one needle cannula, the needle cannula retainer rotating the needle cannula about an axis extending from a first end of the needle cannula to a second end of the needle cannula; an image capture device configured to capture images of the needle cannula at a plurality of positions about the circumference of the needle cannula.

2. The inspection system of claim 1, wherein the needle cannula retainer comprises a wheel having notches along a circumference of the wheel, each notch retaining a needle cannula.

3. The inspection system of claim 2, wherein the needle cannula retainer further comprises a belt tangent to the wheel and contacting the needle cannula to cause rotation of the needle cannula.

4. The inspection system of claim 3, wherein the belt is disposed below a top of a tooth retaining the needle cannula and contacts the needle cannula.

5. The inspection system of claim 1, wherein the control unit is configured to: select an image from the plurality of images with a largest bright area; perform continuous area extraction, contour extraction, curve fitting and ellipse fitting to obtain ideal inner contour, outer contour and burrs; disqualify the needle cannula when a dimension of the burr dimension exceeds a preset threshold; and perform linear fitting of the upper portion of the bright area's outer contour, the two straight lines form an intersection above the needle tip, if the distance between the said intersection and the uppermost point of the bright area exceeds the preset threshold, the needle cannula is disqualified.

6. The inspection system of claim 1, wherein the control unit is configured to: analyze images of the plurality of images captured on one bevel area of said needle cannula; identify an image with a largest bright area; count forward or backward a predetermined number of images to obtain a second image, the second image being approximately 90 degrees to the image with the largest bright area; perform linear fitting of left and right vertical lines; determine the needle has a curved tip when the distance between a top point of the longest vertical line and horizontally the leftmost pixel or rightmost pixel exceeds a preset threshold; and reject the needle cannula.

7. The inspection system of claim 1, further comprising a light source directed at the needle cannula.

8. The inspection system of claim 1, wherein the control unit is configured to detect burrs and a curved tip of the needle cannula form the plurality of images.

9. A machine vision based automatic needle cannula inspection method comprising: rotating a needle cannula about an axis extending from a first end of the needle cannula to a second end of the needle cannula capturing, with an image capture device, a plurality of images of the needle cannula along a plurality of positions about a circumference of the needle cannula; and inspecting the needle cannula for defects through analyzing, by a control unit, the captured images.

10. The needle cannula inspection method according to claim 9, further comprising: adjusting the image capture device and light sources based on a dimension of the needle cannula and at least one of distance, aperture and exposure time to capture the plurality of images; setting, by the control unit, a threshold for each inspection parameter; and rejecting, by the control unit, the needle cannula when a threshold for at least one inspection parameter is not met.

11. The needle cannula inspection method according to claim 9, wherein multithreading techniques and double image capture devices are used to simultaneously inspect bevel areas at each end of the needle cannula.

12. The needle cannula inspection method according to claim 9, further comprising: rotating the needle cannula on a cannula wheel having teeth or a linear stepper belt step having teeth; and rotating the cannula wheel one tooth at a time, or the linear stepper belt step one tooth at a time.

13. The needle cannula inspection method according to claim 9, further comprising processing, by the control unit, the plurality of images with graying, filtering, binarization and rough edge extraction.

14. The needle cannula inspection method according to claim 9, further comprising rotating the needle cannula with a cannula wheel having teeth or a linear stepper belt step having teeth and removing the needle cannula with compressed air.

15. The needle cannula inspection method according to claim 9, wherein the plurality of images captured on one bevel area of the needle cannula are processed to inspect at least one of the following parameters: burrs on bevels and needle tip sharpness; inward or outward curved needle tips; accuracy of several bevel angles of needle cannula tip area; and length of needle cannula.

16. The needle cannula inspection method according to claim 9, further comprising: choosing, by the control unit, an image from the plurality of images with a largest bright area; performing continuous area extraction, contour extraction, curve fitting and ellipse fitting to obtain ideal inner contour, outer contour and burrs; disqualifying the needle cannula when a dimension of the burr dimension exceeds a preset threshold; performing linear fitting of the upper portion of the bright area's outer contour, the two straight lines form an intersection above the needle tip, if the distance between the said intersection and the uppermost point of the bright area exceeds the preset threshold, the needle cannula is disqualified.

17. The needle cannula inspection method according to claim 9, wherein inspecting the needle cannula for defects through analyzing comprises: analyzing images of the plurality of images captured on one bevel area of said needle cannula comprises the following steps: identifying an image with a largest bright area; counting forward or backward a predetermined number of images to obtain a second image, the second image being approximately 90 degrees to the image with the largest bright area; performing linear fitting of left and right vertical lines; determining the needle has a curved tip when the distance between a top point of the longest vertical line and horizontally the leftmost pixel or rightmost pixel exceeds a preset threshold; and rejecting the needle cannula.

Description

DRAWINGS

[0073] Constituting a part of this application, the accompanying drawings are included to provide a further understanding of the invention, exemplary embodiments of the present invention and the descriptions thereof are used to explain the present invention, and do not constitute improper limitation to the present invention. In the drawings:

[0074] FIG. 1 is the side images of a needle cannula tip area at a full revolution or circle

[0075] FIG. 2 is a schematic view toward the needle tip along the axis of needle cannula

[0076] FIG. 3 is a preferred embodiment of the system

[0077] FIG. 4 is on the relationship among the cannula wheel, needle cannula and rotating belt system

[0078] FIG. 5 is a schematic diagram of the inspection of burrs of a needle cannula tip area

[0079] FIG. 6 is a schematic diagram of the inspection of needle tip sharpness

[0080] FIG. 7 is a schematic side view of the needle cannula

[0081] FIG. 8 is a flow chart of the present invention.

[0082] Components in the drawings are labeled as follows: 1-25, side view of 25 images of a needle cannula tip area at equal intervals and a full revolution or circle; 26, burrs at needle cannula tip area; 27, curved needle cannula tips; 28, needle cannula; 29, cannula wheel; 30, tooth grooves; 31, the rotating belt system; 32, camera A; 33, camera B; 34, light source A; 35, light source B; 36, light source C; 37, light sourced D; 38, rejected part removal device using compressed air; 39, tooth groove A; 40, tooth groove B, 41, cylinder; 42, rotating belt; 43, inner contour of needle cannula tip area; 44, outer contour of the needle cannula tip area; 45, burrs; 46, the intersection point of fitting straight lines; 47, uppermost point of needle cannula tip area; 48, right fitting straight line of cannula; 49, left fitting straight line of cannula; 50, curved needle cannula tip; 51, fitting straight line of slope A; 52, fitting straight line of slope B; 53, enlarged view of the needle cannula tip A; 54, enlarged view of the needle cannula tip B

[0083] The present invention will be described below with reference to a preferred embodiment in conjunction with the drawings.

A PREFERRED EMBODIMENT

[0084] It should be noted that, under the condition of no conflict, embodiments of the present invention and the features of the embodiments can be combined with each other. The present invention will be described below with reference to the drawings and the detailed description in conjunction with the embodiments.

[0085] As shown in FIG. 3, the present invention a machine vision based automatic needle cannula inspection system and method, a number of images of needle cannula tip area are captured continuously at equal intervals for a full circle by rotating the needle cannula and the image capture device relatively, multiple quality and technical parameters are inspected through processing and analyzing the images captured using machine vision technologies. After each needle cannula 28 is automatically discharged into the tooth groove 30 of the cannula wheel 29, the cannula wheel pauses when the root of tooth groove 30 is at the top position, the rotating belt system 31 above cannula wheel 29 automatically presses on to the cannula 28 inside the tooth groove 30 and drives the cannula 28 to rotate, the computer sends a signal to start camera A 32 and camera B 33 to capture a number of images at equal intervals and save to the computer under illumination of light source A 34, light source B 35, light source C 36 and light source D 37 when cannula 28 and camera A and Camera B rotates relatively for a full circle, The images are processed using machine vision technologies and results are recorded. rejected parts will be blown away by compressed air of rejected part removal device 38 at rejected parts removal position, qualified needle cannula will advance with the cannula wheel to next process. Meanwhile, at the completion of the needle cannula 28 rotates relatively with camera A 32 and camera B 33 for a full circle, the cannula wheel 29 step advance one tooth after receiving a control signal from the computer indicating the image capture process is completed, awaiting camera A 32 and camera B 33 to receive a signal from the computer to start image capture on next cannula.

[0086] As shown in FIG. 4, tooth groove A and tooth groove B are on cannula wheel 29 to position needle cannula 28, cylinder 41 is sandwiched by the tooth groove A and tooth groove B, the outer diameter of cylinder 41 is on the circumference of the bottom of tooth grooves A and B, tooth groove A 39 and tooth grooves B 40 are symmetrically set to ensure that needle cannula is not bent when pressed and driven to rotate by the power-driven rotating belt system. The width of the power-driven rotating belt is slightly less than the inner distance between the tooth groove A and tooth groove B.

[0087] When a number of images captured on one bevel area of said needle cannula are processed to inspect at least one of the following quality and/or technical parameters: [0088] (1) Burrs on bevels and needle tip sharpness; [0089] (2) Inward or outward curved needle tips; [0090] (3) Accuracy of several bevel angles of needle cannula tip area; [0091] (4) Length of needle cannula.

[0092] The present invention a machine vision based automatic needle cannula inspection method according to a preferred embodiment, comprises the following steps: [0093] (1) Select and adjust said image capture device(s), lenses and light sources based on the dimension of said needle cannula and set parameters such as distance, aperture and exposure time to capture clear images; [0094] (2) Set the accuracy and threshold for each inspection parameter; [0095] (3) Connect the computer with camera A 32 and camera B 33, motors of cannula wheel 29 and rejected part removal unit 38 respectively with synchronized control setting; [0096] (4) Cannula wheel 29 step rotates forward, one tooth at a time, each tooth groove 30 of cannula wheel 29 is automatically placed with one needle cannula, cannula wheel 29 pauses when the bottom of tooth groove 30 is at the upmost position, the rotating belt system 31 above the cannula wheel 29 automatically presses onto the needle cannula 28 in the tooth grove 30 and drive the needle cannula 28 to rotate, a signal sent from the computer starts camera A 32 and camera B 33, under illumination of light source A 34, light source B 35, light source C 36, light source D 37. a number of images are captured at equal intervals while camera A 32 and camera B 33 rotate relatively with needle cannula 28 at a constant speed for a whole circle, the images are saved to the computer and processed with machine vision techniques; [0097] (5) Conventional graying, filtering, binarization and rough edge extraction are processed on saved images. [0098] (6) As shown in FIGS. 5 and 6, when the system performs inspections on needle burrs on bevels and tip sharpness, first choose the one image with largest bright area among all the images captured around a full needle cannula circumference on one tip area, this is the image captured closest to the position when the entire bevel area of the needle cannula is exposed to the camera the most, i.e. position 1 in FIG. 1, then perform accurate continuous area extraction, contour extraction, curve fitting and ellipse fitting to obtain ideal inner contour 43, outer contour 44 and burrs 45, as shown in FIG. 5. When the burr's dimension, if any, exceeds the preset threshold, the needle cannula is disqualified. Then perform linear fitting of the upper portion of the bright area's outer contour, the two straight lines form an intersection 46 above the needle tip, if the distance between the intersection 46 and the uppermost point of the bright area 47 exceeds the preset threshold, the needle cannula is disqualified, as shown in FIG. 6. [0099] (7) As shown in FIG. 7, when the system performs inspections on inward or outward curved needle tips, first identify the image with largest bright area as in step (6), then count forward or backward a number of images, the number is nearest to one-forth of total number of images, in other wards, choose the image that is approximately 90 or 270 degrees to the image with largest bright area, this image is almost the side view of the needle cannula, equivalent to position 7 and 18 in FIG. 1. Then perform linear fitting of the right and left vertical lines 48 and 49 to confirm the longest vertical line 49, if the distance between the top point of the longest vertical line 49 and horizontally the leftmost pixel or rightmost pixel exceeds the preset threshold, the needle cannula has a curved tip and is disqualified. [0100] (8) As shown in FIG. 7, when the system performs inspections on accuracy of several bevel angles of needle cannula, use the same method as in step (7), perform linear fitting on a few lines in the bevel area, in this embodiment there are 3 bevels, so there are 2 fitting lines, i.e., fitting line 51 for bevel A and fitting line 52 for bevel B, angles formed between the cannula left outer line 49 and fitted lines can be calculated and compared with preset threshold to decide whether the needle cannula is qualified. By processing the image of near 180 degrees to this image, the accuracy of bevel angles on the other side can also be inspected. [0101] (9) If the needle cannula inspected has bevel tip areas at both ends, the same steps (5) to (8) can be used to inspect the other end simultaneously. The length of needle cannula can be calculated when both ends of the needle cannula are inspected and the distance between camera A and camera B is calibrated, thus the needle cannula length can be inspected by comparing with preset threshold. When only one end of the needle cannula is inspected, the other end is in a fixed position and the distance between camera and the other end is calibrated, the needle cannula length can also be inspected by comparing with preset threshold. [0102] (10) The system can automatically record classification and statistics of passed and rejected cannulae for query, rejected needle cannula is blown off by compressed air of the rejected part removal device 38 at rejected part removal position while qualified needle cannulae will advance with cannula wheel into the next process. Meanwhile, upon completion of needle cannula 28 rotating relative to camera A 32 and camera B 33 for a full circle and image capture process, cannula wheel 29 controlled by the computer signal advances one tooth and wait for the computer signal to start camera A 32 and camera B 33 to capture images on next needle cannula.

[0103] Besides, images captured along the circumferential direction of a needle cannula for a whole circle can be used to build a three-dimensional model, with which quality and technical parameters can be inspected.

[0104] The descriptions above are only preferable embodiments of the present invention, and are not intended to limit the present invention. To one skilled in the field, the present invention may have various changes and variations. Within the spirit and principle of the present invention, all modifications, equivalent replacements, improvements, etc. are intended to be included within the scope of the present invention.