GB2248934A - Automatic windowing for article recognition - Google Patents

Automatic windowing for article recognition Download PDF

Info

Publication number
GB2248934A
GB2248934A GB9119780A GB9119780A GB2248934A GB 2248934 A GB2248934 A GB 2248934A GB 9119780 A GB9119780 A GB 9119780A GB 9119780 A GB9119780 A GB 9119780A GB 2248934 A GB2248934 A GB 2248934A
Authority
GB
United Kingdom
Prior art keywords
article
orientation
data
work
scan
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
GB9119780A
Other versions
GB2248934B (en
GB9119780D0 (en
Inventor
Randy K Baird
Stanley P Turcheck
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
FMC Corp
Original Assignee
FMC Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US07/583,117 external-priority patent/US5233328A/en
Priority claimed from US07/583,256 external-priority patent/US5103304A/en
Priority claimed from US07/586,189 external-priority patent/US5142591A/en
Priority claimed from US07/586,167 external-priority patent/US5157486A/en
Application filed by FMC Corp filed Critical FMC Corp
Publication of GB9119780D0 publication Critical patent/GB9119780D0/en
Publication of GB2248934A publication Critical patent/GB2248934A/en
Application granted granted Critical
Publication of GB2248934B publication Critical patent/GB2248934B/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01BMEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00Measuring arrangements characterised by the use of optical techniques
    • G01B11/02Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness
    • G01B11/024Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness by means of diode-array scanning
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01BMEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00Measuring arrangements characterised by the use of optical techniques
    • G01B11/02Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness
    • G01B11/028Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness by measuring lateral position of a boundary of the object
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00Programme-control systems
    • G05B19/02Programme-control systems electric
    • G05B19/418Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
    • G05B19/41875Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by quality surveillance of production
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F7/00Methods or arrangements for processing data by operating upon the order or content of the data handled
    • G06F7/02Comparing digital values
    • G06F7/026Magnitude comparison, i.e. determining the relative order of operands based on their numerical value, e.g. window comparator
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/70Determining position or orientation of objects or cameras
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T9/00Image coding
    • G06T9/20Contour coding, e.g. using detection of edges
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/40Extraction of image or video features
    • G06V10/50Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
    • G06V10/507Summing image-intensity values; Histogram projection analysis
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02PCLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • General Engineering & Computer Science (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Manufacturing & Machinery (AREA)
  • Quality & Reliability (AREA)
  • Automation & Control Theory (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Image Processing (AREA)
  • Image Analysis (AREA)

Abstract

In an article handling system that functions to make article discrimination-identification determinations, the possible article orientations are stored and compared (Step 116, Fig. 2) to establish maximum pixel difference numbers and identification of the longitudinal window position along the article where such maximum difference occurs (Steps 118, 120 Fig. 2). Each of the possible orientations is compared with all other possible orientations (loops A. B. Fig. 2) so that a small number of windows is identified as part of a setting-up procedure. When operating, the work articles are scanned and only the data at window locations are used to make article orientation identification, thus reducing the time and memory requirements for data processing. The time and memory may be further reduced by compacting each scan by only storing count values relating to edge locations. <IMAGE>

Description

2 2'+ 3) 34 1 AUTOMATIC WINDOWING FOR ARTICLE RECOGNITION This invention
relates to the determination of article configuration in an article handling system, and more particularly relates to a procedure f or locating positions along the length of an article where maximum differences occur depending upon article configuration.
Background
In Turcheck et al, U.S. Patent Number 4,784,493, an apparatus and method are disclosed for recognition of an article and its orientation on a conveyor. To determine orientation of a work article, a number of possible orientations are first recorded as a preliminary procedure in a memory. All of the data stored in the memory for each orientation is then compared with data scanned from the work article as it moves along a conveyor path. Orientation of the work article is determined by matching the compared data.
The time required for making such article orientation determinations restricts the number of articles that can be processed in a unit of time. Such restricion may be the limiting factor of a production line. Prior efforts at reducing data processing time have included human operator participation in manually selecting by use of a computer keyboard, mouse or the like partial areas along the article length which will be examined. This requires a skill level 2 not always present in the work environment and re-selecting procedures must be undertaken each time there is a change in the article being processed.
Summary of Invention
It is an object of this invention to reduce the processing time for determining article configuration by using only a portion of the object for comparison with stored data where the portion is selected automatically without requiring a skilled operator.
Accordingly, the present invention provides a method of checking the configuration of articles passing along an inspection path by moving articles relative to a scanner to sense article geometric data via a number of successive linear scans of each article, comprising the steps of:
during a setting up procedure:- (a) storing article geometric data for a number of predetermined possible article configurations; (b) for pairs of possible configurations, determining at least one corresponding linear scan in where the difference in data stored in step (a) is large; and during a running procedure:- (c) acquiring comparable geometric data from work articles which is compared with the stored geometric data only at said determined linear scan positions; and (d) determining an article configuration mismatch of the predetermined article configurations and the work article configuration if the difference in the compared 3 geometric data is greater than a predetermined value. A match between the work article configuration and a particular one of the predetermined article configurations may be identified only if the difference in the work article geometric data and the stored geometric data of the particular configuration at said determined scan positions is less than a predetermined value.
The stored and work article geometric data may be shape, orientation or other (eg spatial position) data.
The method of the invention may utilise between at least about four percent and less than about ten percent of the total number of linear scans as determined scans. The method may further include the step of identifying at least one window containing the determined linear scans and including not more than about two adjacent linear scans on each side of the determined linear scans.
The article geometric data may comprise image data relating to the article edge points. The image data may be in the form of count values specifying the location of edge points within each linear scan. In this case the count values may correspond to transition points in serial digital data obtained from the scanner.
Thus, the present invention may provide a method for determining the orientation of work articles transported past a sensing station by comparison of profile data obtained by sequentially scanning each article at spaced positions along the article length as the article moves relative to a scanner 4 to form scan slices with each of said scan slices yielding article edge point data, comprising the steps of:
during a setting procedure:- a) supplying to a memory scan slice data accumulated as the article is transported past said sensing station in a number of different orientations thereby to provide learned edge point data; b) determining scan slice numbers where a large difference exists between the learned edge point data in each possible pair of orientations by comparing said learned edge point data for each possible pair of orientations; c) storing said scan slice numbers; and thereafter during a running procedure, making orientation determinations by: - d) moving singulated work articles past said sensing station; e) sequentially scanning each work article to form scan slices yielding work article edge point data; and f) producing output control signals based on a comparison at the stored scan slice number of the work article edge point data and the learned edge point data without comparing all the learned edge point data.
The invention in one particular form provides apparatus for checking the configuration of articles comprising:
a). means for conveying singulated work articles of like size and shape in a plurality of possible orientations along a path adjacent an article sensor responsive to article geometry; b). said article sensor having a plurality of radiation sensing devices in a linear array that are oriented to be illuminated or shadowed depending upon article geometry and sequentially activated to produce a large number of spaced scan slices at predetermined positions along the article as the article moves along said path; C). means for allowing an increased rate of information processing including:
i). means for generating a signal related to the article profile of a reference article in at least first and second different orientations as part of the setting up procedure; ii). means for comparing said signals corresponding to different orientations to generate a scanned slice position signal at a scanned slice where the article edge points of corresponding scanned slices have a major difference that distinguishes the first orientation from the second orientation as part of the setting up procedure; and iii). means to identify and store said scanned slice position where said major difference exists as part of the setting up procedure; d). means for determining in real time as part of the operating procedure the orientation of a work article on said conveyor which includes said means for generating a signal related to a work article profile, means for examining 6 said work article profile signal at said identified scan slice position and means responsive to a comparison of said examined signal and the signal previously stored at said identified scan slice position for producing an output signal.

Claims (20)

Further features of the invention are in the Claims and are described below by way of example and with reference to the drawings wherein:- Fig. 1 is a pictorial view of a conveyor system having an article diverter; Fig. 2" is a flow diagram of a procedure for generating windows; and Fig. 3 is a pictorial view of four possible orientations of the object whose orientation is to be identified. Detailed Description of Preferred Embodiments The present invention will be described as a feature that is adapted for use with the article or part recognition and re-orientation system disclosed in Turcheck et al U.S. Patent No. 4,784,493. The general environment of the reorientor is diagrammatically illustrated in Fig. 1. The reorientor system may generally comprise a frame supported continuous belt 12 entrained around a driver roll 14 and a idler roll 16. Work pieces such as 18, 20, and 22 are similar parts having three different orientations. The simplest form of reorientor is shown in this figure, that bping a stepping motor driven single axis (Y-axis) reorientor having a lower chamber 26 that can be rotated 180 degrees. other orientors 7 including multiple position reorientors are known in the art and may be advantageously used with the present invention. Adjacent the continuous belt 12 at one edge thereof is a fence 28 running the length of the belt but having several breaks therein. On the inbound side of the reorientor means 24 there is a first break in the fence to accommodate a recognition sensor 30 which may be a 16 X 1 array of vertically stacked fiber optic elements connected to 16 individual phototransistors each having a hard wired connection to a vision controller or microprocessor input port 32. An infrared light source 34 composed of dual infrared LEDs adjusted to different angles is directly across the belt from the recognition sensor 30 and provides the necessary illumination to switch the phototransistors related to each of the 16 fiber optic filaments depending upon whether the individual filament is illuminated or shadowed by the work article. Alternatively, the linear array of sensors may comprise a column of CCD units which provide a pixel density of between about 1000 and 4000 pixels per inch (39-157 pixels/mm)and preferably about 2000 pixels per inch (79 pixels/mm) thereby to provide a high resolution sensor. The CCD units are scanned at a frequency between about 1 MHz and 40 MHz and preferably about 10 MHz to produce an analog signal that is digitized and converted to a count value as disclosed in our co-pending applications filed on even date herewith and titled "High Resolution Parts Handling System" and "High Resolution 8 Camera Sensor Having a Linear Pixel Array", the disclosures of which are hereby incorporated by reference. Hardware compaction of data applied to a microprocessor f or article recognition or orientation determination allows for improved image resolution to be obtained while reducing the processing time and memory size requirements. The present invention is equally usable with the systems disclosed in those applications. The second break in fence 28 is provided to accommodate a first infrared thru beam optical switch composed of a receiver 36 and a light source 38. Immediately prior to the entry port of the orientor means 24 there may be optionally positioned at a third break in the fence 28, a second infrared thru beam optical switch means having a receiver 40 and a light source 42. The recognition sensor communicates via a conduit line 48 with a vision controller 44 which in turn is in communication with an orientation controller 46. Vision controller 44 is hard wired to the work article sensors 30 while the orientation controller 36 is wired to the article recognition sensors 36 and 40 and reorientor 24. A signal related to the movement of the conveyor belt is supplied by lead 58 to orientation controller 46. Control and monitoring of the belt speed maybe by shaft encoder 62 which is connected by lead 60 to vision controller 44, since a constant belt speed is important for maintaining image resolution in this embodiment. 9 The identical sample work pieces 18, 20, and 22 chosen for explanatory purposes of the specification are shown in Figs. 1 and 3 and comprise a plastic article having a length of about 3 inches (7.6 cm) provided with a blunt end surface which may be either at the trailing end as shown at 18 in Fig. 1 to provide orientation A or the leading end as is the case for work article 20 to provide orientation B. The work article 22 is shown with a third orientation C. Up to seven orientations may be determined by the program described below. In operation, work articles 18, 20, and 22 moving along the path of the conveyor belt 12 may be inspected for conformity with a desired and acceptable work piece. In conjunction with such inspection, it is necessary to identify article orientation and make such position changes as are necessary so that all work articles leave the discharge side of the reorientor 24 with the same orientation. Memory resident in the programmable vision controller 44 is "taught" a plurality of up to seven possible orientations of a work article in a setting procedure prior to the production run. The present embodiment is especially adapted for reducing the time required for making the determination of the actual orientation of work pieces, or article identification as the case may be. As explained in the 1493 patent, the capacity for data storage in the vision controller 44 is sufficient to store information concerning the edge points of an article as it passes scanner 30. The standard recognition device operates in a silhouette mode so that only prof ile data is needed. Each scan represents a slice of the article and produces at least one edge point on the profile. The number of slices per article, for example, may be 1000 depending upon conveyor speed, article length and microprocessor programming. An article having acceptable dimensions is fed by the conveyor past the array 30 in a first orientation A. This information is stored in a "learn" mode. Typically this procedure is repeated at least once and optionally up to about ten (10) times to obtain an envelope of values or average value for the first orientation. Next the system is taught to recognize a second orientation B of the same article by the same procedure. Additional orientations C, D.... of the same article up to a total of seven (7) different orientations can be processed by the system of the prior 1493 patent. When all -of the required orientations are taught, i.e. stored in vision controller memory 44, the system is advanced from.the "learn" mode to the "windows generation" mode before moving on to an "operation" mode allowing the repetitive feeding of work articles. Since the conveyor belt speed is carefully controlled, once the article leading edge has been detected, edge point data for corresponding points that are acquired by successive slice scanning can be identified by slices numbered between one and 1000 in the illustrated example. The edge point data are compared to determine which of the orientation data matches the work article data. 11 since the time required f or processing the edge point data has been a factor limiting the speed at which the conveyor 12 may operate, various efforts have been made in the past to reduce the processing time to allow faster classification of objects by the computer. one previous approach has been to have the operator manually set areas of interest with a keyboard, a mouse, or the like. By the present invention, the computer automatically locates the areas of maximum variability between the stored object data and the collected work article data without the need for operator participation. Reference is made to Fig. 2 which shows a flow chart for generating the windows that correspond to numbered slice scans for a specific article whose orientation is to be determined. The procedure illustrated in Fig. 2 will be described in connection with an article that may have four orientations that must be separately ascertained. The program is capable of detecting up to seven orientations as described above. The four orientations A, B, C, and D are shown in Fig. 3. Before starting the program, the orientations are stored just as described in the prior 1493 patent. With the use of the program of Fig. 2, the scan slices 2 - 999 where maximum deviation between the marginal edges that are presented in the several orientations are identified. The article has an arbitrary length of 1000 scan slices that are oriented along the X axis. The article height is arbitrarily designated to be 400 along the Y axis. The thickness of the 12 parts of the article is assumed to be 100 units as measured along the Y axis. With continued reference to Fig. 2, the process is initialized by setting a first loop counter A to zero at step 102. At step 104, the counter is incremented. At steps 106 and 108, the leading edge and the trailing edge scan slices f or orientation A corresponding to X axis positions of 1 and 1000 in Fig. 3 are stored. This corresponds to scan slices 1 and 1000 assuming that a three inch article will be sequentially scanned a thousand times as it passes the sensor 30 of Fig. 1. In this embodiment, scan slices 1 and 1000 are always stored for each orientation. At step 112, a second loop counter B is set to the value of counter A and incremented at step 114 to a position for an iteration with respect to orientation A data. Iteration compares learned data of orientation A with learned data of orientation B by starting with scan slice 2 of both orientation A and orientation B data. The difference in this pixel data at scan slice 2 for orientation A is determined and is called a score. The same procedure is followed for scan slices 3 through 999. In all, 998 scores are determined at step 116. At step 118 a scan slice having a large dif f erence, f or example the maximum score, is determined. While a reading from only one window is theoretically sufficient to determine that a part orientation does not match a stored known orientation, in practice several slice numbers, for example up 13 to about 20, may be stored where the scores are the largest to reduce the likelihood of ambiguity in the results. Thus, at step 126, a determination is made as to whether a sufficient number of windows has been generated. If not, the same procedure is repeated. If "yes", the procedure advances to step 122. At step 122 it is determined whether count B equals the number of learned orientations. If "no" loop counter B is incremented at step 114 to compare the -next orientation data with those of orientation A, after which the slice numbers for the next windows are generated. The loop counter B at step 114 continues incrementing until B is equal to the number of orientations stored. When this condition is satisfied, the procedure passes to step 130. If A is not yet equal to one less than the total number of orientations, A is incremented at step 104. When A equals one less than the total number of orientations at step 130, window determination is complete and the procedure halts at step 134. Thus, it will be seen that counter A iterates from 1 to one less than the total number of learned orientations, whilst for each value of counter A, counter B iterates f rom A + 1 to the total number of learned orientations. In this way, the comparison of like or mirror image data sets is avoided, so avoiding duplicate or null results. At the end of the setting-up procedure, at least 12 scan slice numbers will be identified as windows since each of the four article orientations will have at least three maximum 14 differences (inclusive of the end slices) which each produce three windows. Some of the windows appear at the same scan slice. Where non-standard articles, f or example those having unacceptable tolerances or shapes are to be identified, appropriate windows can be determined automatically by passing one or more acceptable articles through the scanner in the acceptable orientation to generate learned data eg as a part of the foregoing procedure. Then, as a further part of the setting-up procedure one or more sample non-standard articles can be passed through the scanner to locate scan slice positions where the sensed non-standard article data dif f ers most f rom the learned data. These positions will def ine windows appropriate f or sensing non- standard but correctly oriented articles, based on the sample non- standard articles used during the setting-up procedure. Turning to Fig. 3, windows are established by the program of Fig. 2 without operator selection at counts 99, 199, 799 and 899. These are the windows of importance f or article orientation determination in the specific example here being described. - once the windows have been identified, it has been found useful to expand each window to have a width of three or five scan slices centered about the scan slice. Thus, widening of a window compensates for possible data misalignments which can occur in some systems due to mechanical wear and other changes which occur during a continuous operation over several weeks. Af ter the windows are generated as a setting operation, work articles are fed past scanner 30 to identify edge points on the article profile. A comparison operates in real time to determine article orientation during an interval that corresponds to the interval between successive work articles on the conveyor. When a work article is moved past sensor 30 of Fig. 1 which has an orientation A as shown in Fig. 3, a comparison of the work article profile data with each of the learned orientations A, B, C and D is made at the windows previously selected by the program of Fig. 2. Comparing the work article orientation A profile data with stored orientation A data gives a total score of zero. A similar comparison of the same work article data with the stored orientation B data gives a score of 300 at each of the four windows 99, 199, 799 and 899 to thereby produce a total score of 1200. The same comparison with the stored orientation C data gives a total score of 400 and with the orientation D data gives a total score of 1000. From Fig. 3 it can be seen that regardless of which orientation the work article assumes, one stored orientation match with a score at or near zero will be obtained and the orientation of the work article thereby recognized. Where each window is three or five scan slices wide, the score for orientation mismatches increases while remaining essentially at zero for the actual orientation. The results obtained with a comparison of only four or up to about twenty windows along 16 the length of a three inch article can be accomplished with less memory and less time than where all pixel information is processed while at the same time the performance is fully as reliable. Where the high resolution system disclosed in the companion co-pending applications is used, even greater speeds can be obtained which can allow a greater number of articles to be processed per unit time. While only a single embodiment has been illustrated, it is apparent that changes and modifications will be apparent to those skilled in this art. All such modifications and changes which fall within the scope of the claims and equivalents thereof are intended to be covered thereby. 17 CLAIMS:
1. A method of checking the configuration of articles passing along an inspection path by moving articles relative to a scanner to sense article geometric data via a number of successive linear scans of each article, comprising the steps of:
during a setting up procedure:- (a) storing article geometric data for a number of predetermined possible article configurations; (b) for pairs of possible configurations, determining at least one corresponding linear scan in where the difference in data:tored in step (a) is large; (c) acquiring comparable geometric data from work articles which is compared with the stored geometric data only at said determined linear scan positions; and (d) determining an article configuration mismatch of the predetermined article configurations and the work article configuration if the difference in the compared geometric data is greater than a predetermined value.
2. A method of checking article configuration as claimed in Claim 1 wherein the stored and workarticle geometric data comprises article orientation data.
is
3. A method of checking article configuration as claimed in Claim 1 or Claim 2 wherein the stored and work article geometric data comprises article shape data.
4. A method of checking article configuration as claimed in any preceding Claim, further comprising the step of identifying a match between the work article configuration and a particular one of the predetermined article configurations only if the difference in the work article geometric data and the stored geometric data of the particular configuration at said determined scan positions is less than a predetermined value.
5. A method of checking article configuration as claimed in Claim 4, further comprising the step of identifying a plurality of said determined scan positions; totalling the sum of the differences for said plurality of determined linear scans and using said sum differences for identifying work article configuration.
6. A method of checking article configuration as claimed in any preceding Claim, wherein the number of said determined linear scans is at least about four percent and less than about ten percent of the total number of linear scans.
7. A method of checking article configurations as claimed in any preceding Claim, including a further step during the setting procedure of identifying at least one window containing said determined linear scans and including nor more than about two adjacent linear scans on each side of 19 the determined linear scans.
8. A method of checking article configuration as claimed in any preceding Claim, wherein the article geometric data comprises image data relating to the article edge points.
9. A method of checking article configuration as claimed in Claim 8, wherein the article image data comprises count values specifying the location of edge points within each linear scan.
10. A method of checking article configuration as claimed in Claim 9, wherein the count values correspond to transition points in serial digital data obtained from the scanner.
11. A method for determining the orientation of work articles transported past a sensing station by comparison of profile data obtained by sequentially scanning each article at spaced positions along the article length as the article moves relative to a scanner to form scan slices with each of said scan slices yielding article edge point data, comprising the steps of:
during a setting procedure:- a). supplying to a memory scan slice data accumulated as the article is transported past said.sensing station in a number of different orientations thereby to provide learned edge point data; b). determining scan slice numbers where a large difference exists between the learned edge point data in each possible pair of orientations by comparing said learned edge point data for each possible pair of orientations; c). storing said scan slice numbers; and thereafter during a running procedure, making orientation determinations by:
d). moving singulated work articles past said sensing station; e). sequentially scanning each work article to form scan slices yielding work article edge point data; and f). producing output control signals based on a comparison at the stored scan slice number of the work article edge point data and the learned edge point data without comparing all the learned edge point data.
12. A method as claimed in Claim 11, wherein at least those scan slice numbers where the maximum difference exists between the learned edge point data in each possible pair of orientations are stored.
13. A method for determining orientation as claimed in Claim 11 or Claim 12 wherein the comparison of the work article and learned edge point data is made by subtracting work article edge point data from learned edge point data only at windows defined by the stored scan slice numbers to produce a total score relating to the windows for each of the possible orientations sequentially, whereby the lowest total score is used to identify the work article orientation.
14. The method of Claim 13, wherein the windows include not more than about two adjacent scan slices on each side of a scan slice whose number is stored.
21
15. Apparatus for checking the configuration of articles comprising: a). means for conveying singulated work articles of like size and shape in a plurality of possible orientations along a path adjacent an article sensor responsive to article geometry; b). said article sensor having a plurality of radiation sensing devices in a linear array that are oriented to be illuminated or shadowed depending upon article geometry and sequentially activated to produce a large number of spaced scan slices at predetermined positions along the article as the article moves along said path; c). means for allowing an increased rate of information processing including: i). means f or generating a signal related to the article prof ile of a ref erence article in at least f irst and second different orientations as part of the setting up procedure; ii). means for comparing said signals corresponding to different orientations to generate a scanned slice position signal at a scanned slice where the article edge points of corresponding scanned slices have a major difference that distinguishes the first orientation from the second orientation as part of the setting up procedure; and, iii). means to identify and store said scanned slice position where said major difference exists as part of the setting up procedure;
22 d). means for determining in real time as part of the operating procedure the orientation of a work article on said conveyor which includes said means for generating a signal related to a work article profile, means for examining said work article profile signal at said identified scan slice position and means responsive to a comparison of said examined signal and the signal previously stored at said identified scan slice position for producing an output signal.
16. Apparatus as defined in Claim 15 wherein the sensor includes a plurality of radiation sensing elements aligned in a linear array and disposed to face a source of radiation so that different ones of said radiation sensing elements are illuminated while others are shadowed thereby to sense only edge points on an article profile.
17. Apparatus as defined in Claim 16 wherein the sensor comprises vertically stacked linear array of fiber optic elements each connected to a respective photo transistor.
18. Apparatus as defined in Claim 16 wherein the sensor comprises vertically stacked linear array of CCD units.
19. Apparatus as defined in Claim 18 further having means for scanning at least about 1000 CCD units for each sequential scan to produce an analog image data output signal and means for compressing the analog image data applied to said comparing means, thereby to reduce the processing time and memory size requirements of the comparing means.
20. Apparatus as defined in Claim 19 wherein data in each of said scanned slices is part of an output signal from a 23 scanner that is synchronized with a counting circuit and a counting circuit count value corresponding to the time of occurrence of a point on said article profile in said information data is supplied to said comparing means through a first in, first out buffer memory.
GB9119780A 1990-09-17 1991-09-16 Automatic windowing for article recognition Expired - Fee Related GB2248934B (en)

Applications Claiming Priority (6)

Application Number Priority Date Filing Date Title
US07/583,117 US5233328A (en) 1990-09-17 1990-09-17 Method for processing compacted data
US07/583,256 US5103304A (en) 1990-09-17 1990-09-17 High-resolution vision system for part inspection
US07/586,189 US5142591A (en) 1990-09-21 1990-09-21 High resolution camera with hardware data compaction
US07/586,167 US5157486A (en) 1990-09-21 1990-09-21 High resolution camera sensor having a linear pixel array
US58693990A 1990-09-24 1990-09-24
US58744890A 1990-09-25 1990-09-25

Publications (3)

Publication Number Publication Date
GB9119780D0 GB9119780D0 (en) 1991-10-30
GB2248934A true GB2248934A (en) 1992-04-22
GB2248934B GB2248934B (en) 1994-11-30

Family

ID=27560152

Family Applications (6)

Application Number Title Priority Date Filing Date
GB9119777A Expired - Fee Related GB2248685B (en) 1990-09-17 1991-09-16 High-resolution vision system for part inspection
GB9119774A Expired - Fee Related GB2248931B (en) 1990-09-17 1991-09-16 High resolution parts handling system
GB9119780A Expired - Fee Related GB2248934B (en) 1990-09-17 1991-09-16 Automatic windowing for article recognition
GB9119776A Expired - Fee Related GB2248933B (en) 1990-09-17 1991-09-16 High resolution camera with hardware data compaction
GB9119775A Expired - Fee Related GB2248932B (en) 1990-09-17 1991-09-16 Method for processing compacted data
GB9119778A Expired - Fee Related GB2248686B (en) 1990-09-17 1991-09-16 High resolution camera sensor having a linear pixel array

Family Applications Before (2)

Application Number Title Priority Date Filing Date
GB9119777A Expired - Fee Related GB2248685B (en) 1990-09-17 1991-09-16 High-resolution vision system for part inspection
GB9119774A Expired - Fee Related GB2248931B (en) 1990-09-17 1991-09-16 High resolution parts handling system

Family Applications After (3)

Application Number Title Priority Date Filing Date
GB9119776A Expired - Fee Related GB2248933B (en) 1990-09-17 1991-09-16 High resolution camera with hardware data compaction
GB9119775A Expired - Fee Related GB2248932B (en) 1990-09-17 1991-09-16 Method for processing compacted data
GB9119778A Expired - Fee Related GB2248686B (en) 1990-09-17 1991-09-16 High resolution camera sensor having a linear pixel array

Country Status (1)

Country Link
GB (6) GB2248685B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105136045A (en) * 2015-09-22 2015-12-09 北京佰能电气技术有限公司 Coil down length measurement method for coil collecting station

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
SE9801170L (en) * 1998-04-02 1999-10-03 Photonic Systems Ab Method and system for monitoring or scanning an object, material or the like
CN108445808A (en) * 2018-03-30 2018-08-24 深圳前海清科技有限公司 The sensing device and method that data synchronize

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0067438A2 (en) * 1981-06-15 1982-12-22 Oxbridge, Inc. Automatic optical inspection apparatus
GB2140603A (en) * 1983-05-27 1984-11-28 Pa Consulting Services Adaptive pattern recognition
US4711579A (en) * 1986-08-12 1987-12-08 H. Fred Johnston System for automatically inspecting a flat workpiece for holes

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4608709A (en) * 1983-03-08 1986-08-26 Owens-Illinois, Inc. Method and apparatus for gauging containers
GB2175396B (en) * 1985-05-22 1989-06-28 Filler Protection Developments Apparatus for examining objects
US4678920A (en) * 1985-06-17 1987-07-07 General Motors Corporation Machine vision method and apparatus

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0067438A2 (en) * 1981-06-15 1982-12-22 Oxbridge, Inc. Automatic optical inspection apparatus
GB2140603A (en) * 1983-05-27 1984-11-28 Pa Consulting Services Adaptive pattern recognition
US4711579A (en) * 1986-08-12 1987-12-08 H. Fred Johnston System for automatically inspecting a flat workpiece for holes

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105136045A (en) * 2015-09-22 2015-12-09 北京佰能电气技术有限公司 Coil down length measurement method for coil collecting station
CN105136045B (en) * 2015-09-22 2018-01-05 北京佰能盈天科技有限公司 One kind collection volume station, which is coiled, surveys long method

Also Published As

Publication number Publication date
GB2248931A (en) 1992-04-22
GB9119774D0 (en) 1991-10-30
GB2248934B (en) 1994-11-30
GB2248933A (en) 1992-04-22
GB2248685A (en) 1992-04-15
GB2248933B (en) 1994-08-31
GB2248685B (en) 1994-10-19
GB9119777D0 (en) 1991-10-30
GB9119776D0 (en) 1991-10-30
GB2248932B (en) 1994-10-12
GB9119775D0 (en) 1991-10-30
GB9119778D0 (en) 1991-10-30
GB2248931B (en) 1995-01-04
GB2248686B (en) 1994-12-14
GB9119780D0 (en) 1991-10-30
GB2248686A (en) 1992-04-15
GB2248932A (en) 1992-04-22

Similar Documents

Publication Publication Date Title
EP1472165B1 (en) Apparatus and method for detecting overlapped mail pieces
US20100260378A1 (en) System and method for detecting the contour of an object on a moving conveyor belt
US5106195A (en) Product discrimination system and method therefor
EP1363228B1 (en) Method and machine for reading and associating optical codes
US4678920A (en) Machine vision method and apparatus
US5311977A (en) High resolution parts handling system
JP3041090B2 (en) Appearance inspection device
US10189055B2 (en) Color based optical grading system with multi reflectance and multi-angle views
US5353358A (en) Method for determining article orientation in an article handling system using a sampling of a CCD array
Yang Finding stalk and calyx of apples using structured lighting
US4784493A (en) Element recognition and orientation
EP0917649A1 (en) Apparatus and method for detecting surface defects
CN211070921U (en) Instrument appearance detection device based on 3D scanning method
Conners et al. Machine vision technology for the forest products industry
US20070071284A1 (en) Method of detecting bunched-together poster items by analyzing images of their edges
GB2248934A (en) Automatic windowing for article recognition
US5018864A (en) Product discrimination system and method therefor
CN118243220A (en) Method and device for detecting the pitch/elevation of a fish transported by a conveyor
CN111687061A (en) Automatic detection equipment for lens imaging sensor
WO2011124583A1 (en) Method and device for the controlled transport of multiple objects
US5764788A (en) Strand orientation sensing
DK155274B (en) DEVICE FOR CONTROL OF WOODEN TOPICS
JPH11271024A (en) Measuring apparatus of fruit and vegetable
CA1281392C (en) Lumber optimizer
JPH11271025A (en) Measuring apparatus of fruit and vegetable

Legal Events

Date Code Title Description
PCNP Patent ceased through non-payment of renewal fee

Effective date: 20020916