EP0394348A1 - Procede et appareil d'evaluation optique d'objets inconnus - Google Patents
Procede et appareil d'evaluation optique d'objets inconnusInfo
- Publication number
- EP0394348A1 EP0394348A1 EP89901660A EP89901660A EP0394348A1 EP 0394348 A1 EP0394348 A1 EP 0394348A1 EP 89901660 A EP89901660 A EP 89901660A EP 89901660 A EP89901660 A EP 89901660A EP 0394348 A1 EP0394348 A1 EP 0394348A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- signal data
- transform image
- generating
- signature
- defining
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
- G01N21/956—Inspecting patterns on the surface of objects
- G01N21/95623—Inspecting patterns on the surface of objects using a spatial filtering method
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/88—Image or video recognition using optical means, e.g. reference filters, holographic masks, frequency domain filters or spatial domain filters
Definitions
- This invention relates to an optical object inspection system, and more specifically relates to the method and apparatus by which the optical inspection system evaluates the conformance of an unknown object to predetermined characteristics.
- a characteristic signature of each object undergoing inspection based upon its transform image, and this characteristic signature is compared to a predetermined characteristic signature.
- the transform image may be a Fourier transform image
- the characteristic signatures for the known and unknown objects may be generated using the same equipment and methods.
- Background of the Invention Machine vision or inspection systems have become a vital component in integrated manufacturing systems. They can sort, package, and perform defect analysis without human intervention. For instance, by inspecting holes being drilled the system can determine if a drill bit is worn.
- the image to be processed is converted into a Fourier or other transform.
- the Fourier transform presents information about the image of the object in a very useful, symmetrical pattern which represents the object in terms of its spatial frequencies.
- the calculation of a Fourier transform on a digital computer is extremely intense, and may take a computer as powerful as a Micro Vax II about a minute to complete. Even powerful and expensive state of the art array processors take a full second to merely produce the transform. In modern industrial plants, the production line rates are often a full order of magnitude faster than this.
- the computational intensity and time are significantly reduced using parallel processing tech ⁇ niques, such as those available when the real image of the object undergoing inspection is converted to a trans- form image and optically processed.
- parallel processing tech ⁇ niques such as those available when the real image of the object undergoing inspection is converted to a trans- form image and optically processed.
- Following generation of the transform image it is "processed” by quantifying the light from a preselected number of spatial domains or segments of the transform image. These quantities are then electronically processed to provide an abbreviated or composite characteristic signature of the transform image, and thus of the object upon which it is based.
- the signatures may be rapidly and economically obtained and then evaluated to determine whether the object does or does not conform to preselected standards.
- an inspection system of the foregoing type must include signature-generating means, in addition to the other system components such as means for generating electrical signal data representative of the appearance of each inspected object, means for receiving such signal data and producing a visual image of each object represented thereby, and means for producing a transform image of the object from the aforesaid visual image.
- signature-generating means in addition to the other system components such as means for generating electrical signal data representative of the appearance of each inspected object, means for receiving such signal data and producing a visual image of each object represented thereby, and means for producing a transform image of the object from the aforesaid visual image.
- the present invention provides an improved method and means for optically evaluating the conformance of unknown objects to predetermined charactertistics and includes the steps of generating a transform image of a known object; sampling the light from different spatial domains of the transform image and generating signal data representative of the different domains; collecting the signal data and collectively defining and storing a characteristic electric signature vector representing the transform image of the known object; generating an optical trans ⁇ form of an unknown object; sampling light from different spatial domains of the transform image of the unknown object and generating signal data representative of the different domains; collecting the signal data and collec ⁇ tively defining a characteristic electric signature vec ⁇ tor representing a transform image of the unknown object; and comparing the electric signature vectors presentating the unknown and known objects and determining whether the signature vector of the unknown object conforms to the boundaries of the signature vector for the known object.
- the method includes sampling the light intensity as one method of characterizing the different domains of the transform image,, generating optical transform images of a variety of known objects to define a range of values for each domain of the transform optical image, and deter- mining statistical variations among certain ones of the sampled light quantities from the different spatial domains.
- the apparatus includes first imaging means for generating electrical signal data representative of the appearance of each inspected object, second imaging means for receiving the signal data and generating a visual image of the object, third imaging means for receiving and generating a transform image from the visual image of the object, and signature generating means for sampling light from different spatial domains of the transform image and generating electrical signal data represen ⁇ tative thereof and collectively defining a characteristic signature of the inspected object.
- a control means controls and coordinates the operation of the foregoing system components so as to permit both quite rapid and highly accurate inspection of the objects.
- the signature generating means of the inspection system includes movable mask means for obtaining samples of light from different spa- tial domains of the transform image, a system control means synchronizes the operation of other system com ⁇ ponents and the movement of the mask means, and a com ⁇ puter collects the data representing the intensity of the light from the different spatial domains of the transform image, assembles the data to define a characteristic electric signature vector representing the transform image of the known object, and compares this formed signature vector to a corresponding signature vector representing the transform image of the unknown object.
- the first imaging means of the inspection system may and preferably does include sensing.means for sensing the presence of an object at a preselected inspection- location, stroboscopic lighting means for illuminating each object at such location, and video camera or similar means for generating electrical signals representative of the appearance of an object present at such location.
- the control means initiates operation of the stroboscopic lighting means and thereafter ensuing transmission of the desired signal data from the camera means only after it has verified that the object in question is within the field of view of the camera means, and the latter has completed its transmission of signal data representative of the appearance of a preceding object at the inspection location.
- the control means preferably further delays operation of the light-sampling and other components of the signature generating means of the system until the optical real and transform images of the object undergoing inspection are of an optimal quality.
- the optimal transform image of the object has been generated, the light sampling or mask means iso ⁇ lates different spatial domains of the transform image and a photodetector generates an electrical signal repre ⁇ sentative of the light intensity in each of the different domains.
- the electrical signal data is stored corresponding to each domain and this correlated data collectively defines an electric signature vector repre ⁇ senting the transform image of the known object.
- This same technique is used for collecting the signature of a known object, whether it be a "good” or "bad” object, and the signature of an unknown object.
- the signature vec ⁇ tors may be compared, after being preprocessed as necessary or desirable, to ultimately determine whether the unknown object conforms to the predetermined charac- teristics of the known object.
- the method if more than merely a means of determining conformance of an unknown object to a prede ⁇ termined signature, but also, a method for organizing and generating the signature in the first place.
- the signature allows the global characterization of the complete appearance of the object.
- the method contemplates gathering many signature, clustering and analyzing the signatures, forming a generalization of a correct signature, and setting meaningful thresholds between good and bad.
- Fig. 1 is a schematic and diagrammatic view of operational and control components of an optical object inspection system in accordance with the invention
- Fig. 2 is an enlarged side elevational view of a rotatable mask component of the system
- Fig. 3 is a map or graph of the intensity of the transform image in each of sixteen domains, representing the signature of the object being inspected;
- Fig. 4 is a typical map or graph of the signa ⁇ tures of numerous objects with the multiple signatures being superimposed one on the other to define acceptable ranges for each of the 16 domains.
- Fig. 5 is a schematic flow chart illustrating the method of operation of the present invention using the apparatus disclosed herein and in the parent and related applications.
- the numeral 10 in Fig. 1 designates conveyor means for sequentially conducting objects 12, only one of which is shown, in the direction of the arrow 14 along a path of travel extending past an "upstream" object inspection station 16 and a "downstream” object reject station 18.
- Objects 12 can and normally would be transported by conveyor 10 at a relatively high speed, such as 15 objects per second, and might be packaged, canned, bottled or other items intended to be of a generally uniform or standardized type. To ensure that objects 12 meet preestablished standards of uniformity, each object is automatically inspected as it reaches inspection station 16.
- the system for inspecting objects 12 includes first imaging means for producing electrical signal data representative of the appearance of each object arriving at inspection station 16, second imaging means for receiving the aforesaid electrical signal data and producing a visual image of the object, third imaging means for producing a transform image of the object from the aforesaid visual image, signature generating means for sampling light from a limited number of different spatial domains of the transform image and generating therefrom electrical signal data defining a characteristic signature or summary of the transform image and thus of the inspected object, and control means for automatically correlating and controlling operation of the foregoing components.
- the first imaging means includes sensing means 22, such as a photoelectric or sonic type object detector, for sensing and signalling the arrival of each object 12 at inspection station 16; stroboscopic lighting means 24 for, when actuated, briefly illuminating the object at the inspection station; and video camera means 26 for capturing the real image of the illuminated object at the inspection station, and producing electrical signal data representative of the real image of the object.
- the second imaging means of the inspection system includes frame buffer means 28, and spatial light modulator ("SLM”) means 30 which may be and preferably is a liquid crystal display device.
- SLM spatial light modulator
- Frame buffer 28 receives the signal data essentially in real time as it is generated by camera 26, and retransmits the data to SLM 30 at a predetermined time, causing the SLM to pro ⁇ turn a high-quality visual image of the inspected object 12.
- the frame buffer is typically a fast access random access memory device that stores each line of real image data as it is generated. When all of the real image data has been stored, i.e. the last horizontal trace has been generated and stored, all of the data is transmitted to the spatial light modulator. And, when caused to do so by the control means of the inspection system, frame buffer 28 also causes rapid quenching of the image generated by SLM 30.
- the third imaging means of the inspection system includes a laser 32 or other means for producing a coherent beam of light that passes through SLM 30 and defines the optical axis of the system. In passing through the SLM the real image generated by the SLM is impressed on the beam of light.
- the third imaging means further includes lens means 34 that receives the now modulated light beam carrying the real image of the object 12, and converts that image into a Fourier or other transform image of the inspected object.
- the sampling and signature generating means of the apparatus include movable mask means, illustratively in the form of a rotatable disc 36, and light detecting and measuring means such as a photodetector device 38.
- Disc 36 is adapted to be rotated at a constant speed about its central axis by a servo motor 40, and is also shown in Fig. 2 of the drawings.
- Disc 36 has a circular array of wedge and ring shaped masks 42 of differing orientations, configurations and/or sizes. Each mask may represent a different spatial domain or segment of the transform image, but together they represent a composite or mosaic of entire image.
- the disc 36 includes thirty two masks, divided into two sets of sixteen.
- One set of sixteen includes 16 wedge or "bow tie" shaped masks, and the other set includes 16 ring or donut shaped masks.
- each of the donut shaped masks has a different radius so that together they also form a composite of the entire image area.
- the disk also includes timing marks 41 at its perimeter. The timing marks may be masks and may be optically detected using one of several known techniques or sensing means 46 to determine the speed and precise angular orientation of the disc.
- the signal data generated by device 38 repre ⁇ senting the intensity of the transform image domains defined by each of the masks 42 is processed by conven ⁇ tions signal processing means 43.
- This may include such things as a current-to-voltage converter, a sample and hold circuit, an analog-to-digital converter, and a buffer means.
- the output data from signal processing means 43 is received by a computer, which is also part of the control means 44 of the inspection system.
- the data received by the computer represents the light intensity for each domain of the transform image. By mapping the intensity for each domain, as shown in Figure 3, a unique signature line S is generated for each object inspected.
- the computer contains data, which may be based upon prior analysis of the characteristic signatures of a number of "good” objects previously inspected by the system, that is used to determine from its characteristic signature data whether the object 12 currently undergoing inspec ⁇ tion falls within the acceptable limits defining a "good” or "bad” category.
- controller 44 directs a control signal to pusher mechanism 20 causing the object 12 in question to be displaced by the pusher mechanism from conveyor 10 when the object reaches reject station 18. Controller 44 may also maintain a record of the number of bad objects detected during a particular time period and actuate an alarm 47 if such number reaches a predetermined limit. In addition to the foregoing functions, controller 44 automatically controls and coordinates the operation of various of the components of each of the three imaging means and the signature generating means of the inspection system. At the outset of a typical cycle of operation of the inspection system, the arrival of an object 12 at inspection station 16 and within the field of view of camera 26 is detected by detector 22 and reported to controller 44.
- Controller 44 continuously monitors the operation of camera 26, display device or SLM 28 and photodetector 38. If the condition of the latter com- ponents is suitable for commencement of a cycle of opera ⁇ tion of the inspection system, the controller 44 initiates the same upon receipt of the signal from detec ⁇ tor 22. If one or more of the aforesaid components is not in suitable condition, initiation of the inspection cycle is delayed until it or they are in proper con ⁇ dition.
- controller 44 When the object is in the proper position, controller 44 effects actuation of strobe light 24 and transmission from camera 26 to frame buffer 28 of signal data representative of the real image of the inspected object "12. Such data is received and stored by frame buffer 28, and when the entire real image has been stored it is then transmitted to the liquid crystal SLM 30.
- the SLM may thereafter be refreshed by the frame buffer for a predetermined time period as necessary and specified by controller 44 and commensurate wtih the response time of the SLM 30 employed in the inspection system.
- the visual image produced by SLM 30, and then impressed upon the light beam generated by laser 32, is converted by lens means 34 into a transform image.
- the different spatial domains are viewed by photodetector 38 as different ones of the mask apertures 42 of disk 36 move, as the disk rotates, into the optical path of the detector 38.
- Constant-speed rotation of disk 36 by its servo motor 40 is initiated by controller 44 when the inspec ⁇ tion system is first placed into operation, and is thereafter monitored by the controller.
- the timing marks 41 present upon the disk 36 are continuously scanned by sensing means 46 (Fig. 1) , of a tachometer type, asso ⁇ ciated with the disk.
- the output of tachometer 46 con ⁇ tinuously produces a signal representative of the uniform rate of rotation of the disk.
- This signal also permits identification by controller 44 of the instantaneous rotative position of respective ones of the disc aper ⁇ tures 42 relative to the optical path between photodetec ⁇ tor 38 and the transform image.
- the tachometer signal is utilized by the controller 44 to correlate the light intensity signal data generated by photodetector 38 to each successive mask aperture 42 (i.e. transform image spatial domain) of the disk. Such correlation is necessary for proper identification and subsequent com ⁇ parison, by the computer component of controller 44, of the characteristic signature signal data generated during each inspection cycle with the characteristic signature signal data stored in the computer and relating to "good" objects.
- the acceptance of signals from the photodetector 38 during each inspection cycle is delayed until the ima- ges generated by devices 30, 34 are in an optimum con ⁇ dition.
- the duration of the delay depends upon the response time of the particular device 30 employed in the inspection system, as well as possibly other factors.
- the transform image was most stable during the last 24 milliseconds of each cycle, and the charac ⁇ teristic signature of the image was obtained during such terminal period. It is of course understood that during such period the generation and/or transmission of signal data by detector 38 is permitted by controller 44 only when successive ones of the mask apertures 42 are aligned with the optical path between the transform image and device 38.
- the signatures which represent the distribution of spatial frequencies in the transform image are collected and stored in a sample file.
- the light inten ⁇ sity for each domain of the transform image represents a feature of the signature.
- the line drawn between the intensity points for each domain, as illustrated in Figure 3, represents the signature S itself.
- the corresponding features of all signatures in the sample file are averaged to form a composite signature for each sample file.
- the difference between the highest and lowest intensities for each domain or feature of the signature represents the range within which a domain intensity or feature must fall to be acceptable.
- the average signature for a group of similar products is designated by reference numeral A and the acceptable range of intensity for domains 1 is represented by reference numeral R.
- the signature may be comprised of selected subsets of data from the sampled transform image.
- a "shape" signature typically represents the orientations of the transform image frequencies or the shape of the image as generated by the wedge shaped masks.
- a "texture” signature represents the frequency distribution of the transform image or its texture as generated by the ring shaped masks.
- a further technique utilizes an "offset" which causes the apparatus to subtract any system background signature that may be included within a given signature.
- a group of good objects is passed in front of the camera and their signatures are recorded to generate a sample or average signature vector repre ⁇ senting the composite transform image of the good objects.
- statistical evaluations may be generated to evaluate the validity of the collected data or the significance of variations for one or more of the domains. This is typically done in terms of distance metrics.
- Another technique is to generate a covariance matrix, which collects numerical data reflecting the manner in which the characteristics in different domains vary in the same diretion and magnitude.
- the Euclidian distance for a signature is calculated by adding the squared differences of each features' value and its mean value for all the features of signatures in a sample file and deriving the square root of this sum.
- the Euclidian distance accumulates separation for every feature of a signature. While there may not be much separation between one feature and another, total separa ⁇ tion of signatures from one sample file to another can be substantial and significant.
- Another evaluation that is useful is to determine the largest separation between features of the signatures. This metric is easy to com- pute and is not sensitive to cumulative differences among features.
- An average is calculated which represents the average of all of the signatures in the sample file.
- a standard deviation is calculated which is a measure of the variation from the mean among the items in the sample.
- a broader range may be established by defining a threshold which is midway between the range of signatures for good samples and the range of signatures for rejected samples. This provides a larger measure of acceptance for good objects, yet ensures a minimum distance from those signatures which characterize reject objects.
- the threshold may be placed automatically or manually, as necessary or desirable. Through these pro- cedures it can sense a clustering of signatures at dif ⁇ ferent points in the metric and place the thresholds automatically to distinguish between the segments of the metric represented by the clustering of the signatures. The larger the distance or window between the signatures representing the good objects and the rejected objects, the higher the probability for accurate discrimination between good and rejected objects during product inspec ⁇ tion.
- the sequence of steps 51 - 54 represent the generation of the optical transform image of a known object, sampling of the light intensity in the transform domains for each of the known objects, and the storing of the intensity for each domain of each 6f the sampled objects.
- This loop continues for the entire group of known objects.
- the electrical signal data representing the intensity for each domain of each known object has'been stored, they are collected to define a characteristic electric signature vector repre ⁇ senting the transform image of the known object.
- This signature vector is typically a compilation of all of the intensities in each domain, and the intensities may be used to establish an average signature vector or a range of values that represent acceptable characteristics of the Fourier transform for the various domains of the object. This is represented by box 55 in Figure 5.
- Steps 56 and 57 represent the generation of an optical transform image of an unknown object and the sampling of light from the different spatial domains of the transform image of the unknown object in the same manner, as was done in connection with the known objects.
- Electrical signal data representative of the sampled light from each of the different domains thereof is collected to define a characteristic electric signature vector representing the transform image of the unknown object.
- the electric signature vector representing the transform image of the unknown object is compared, domain by domain, with the electric signature vector repre ⁇ senting the transform image of the known object, and may be implemented using a neural network or standard com- puter processing techniques.
- this comparison may represent the degree of conformance to the average signature, the number of domains for which the transform of the unknown object matches the average signature for the known object, the number of domains for which the signature of the unknown object falls within the acceptable ranges for the domains of the transform image of the known object, etc.
- Statistical evaluations of the number of good versus reject objects, as well as the highest degree of variation present, may be performed.
- a decision representing the acceptance or rejectance of the inspected object is formulated and an electrical signal representative thereof is transmitted to a downstream station to collect or detour the unknown object.
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Chemical & Material Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
On évalue la conformité d'un objet inconnu (12) à des caractéristiques prédéterminées à l'aide d'une signature caractéristique, en utilisant son image de transformation. On détecte la présence à une station d'inspection (16) de chacun des objets successifs (12) à inspecter. Le système d'inspection comprend un disque rotatif (36) comportant des ouvertures de masque (42) se déplaçant dans la ligne de mire d'un détecteur de lumière (38). Les signaux produits par le balayage de marques de temps (41) situées sur le disque (36) reflètent celui d'une pluralité de masques (42), qui est présent dans la trajectoire optique, et le photodétecteur (38) détermine l'intensité de l'image de transformation échantillonnée par ledit moyen de masque (42). Les signaux représentant ces intensités, mis en corrélation avec le domaine duquel ils émanent, sont collectés afin de former une signature pour ledit objet. On compare cette signature à une signature connue.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US13741787A | 1987-12-23 | 1987-12-23 | |
US137417 | 1987-12-23 |
Publications (2)
Publication Number | Publication Date |
---|---|
EP0394348A1 true EP0394348A1 (fr) | 1990-10-31 |
EP0394348A4 EP0394348A4 (en) | 1992-09-30 |
Family
ID=22477340
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP19890901660 Withdrawn EP0394348A4 (en) | 1987-12-23 | 1988-12-22 | Method and apparatus for optically evaluating unknown objects |
Country Status (4)
Country | Link |
---|---|
EP (1) | EP0394348A4 (fr) |
JP (1) | JPH03501785A (fr) |
BR (1) | BR8807862A (fr) |
WO (1) | WO1989006020A1 (fr) |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0265194A2 (fr) * | 1986-10-17 | 1988-04-27 | Global Holonetics Corporation | Système de traitement d'image utilisant une transformation optique |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS5437459B2 (fr) * | 1972-01-13 | 1979-11-15 | ||
US4139303A (en) * | 1977-01-21 | 1979-02-13 | The Board Of Regents Of The University Of Washington | Adaptive coherent optical processing method and apparatus for recognizing and counting objects |
US4360799A (en) * | 1980-05-22 | 1982-11-23 | Leighty Robert D | Hybrid optical-digital pattern recognition apparatus and method |
-
1988
- 1988-12-22 EP EP19890901660 patent/EP0394348A4/en not_active Withdrawn
- 1988-12-22 BR BR888807862A patent/BR8807862A/pt unknown
- 1988-12-22 JP JP1501610A patent/JPH03501785A/ja active Pending
- 1988-12-22 WO PCT/US1988/004636 patent/WO1989006020A1/fr not_active Application Discontinuation
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0265194A2 (fr) * | 1986-10-17 | 1988-04-27 | Global Holonetics Corporation | Système de traitement d'image utilisant une transformation optique |
Non-Patent Citations (2)
Title |
---|
PROCEEDINGS OF THE SPIE CONFERENCE: OPTICS, ILLUMINATION AND IMAGE SENSING FOR MACHINE VISION II vol. 850, 5 November 1987, CAMBRIDGE, US pages 66 - 80; D.P.CASASENT & J. RICHARDS: 'An optical processor for product inspection' * |
See also references of WO8906020A1 * |
Also Published As
Publication number | Publication date |
---|---|
EP0394348A4 (en) | 1992-09-30 |
JPH03501785A (ja) | 1991-04-18 |
WO1989006020A1 (fr) | 1989-06-29 |
BR8807862A (pt) | 1990-11-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5078501A (en) | Method and apparatus for optically evaluating the conformance of unknown objects to predetermined characteristics | |
CA1313243C (fr) | Methode et moyen de commande pour systeme d'inspection optique | |
US5703960A (en) | Lumber defect scanning including multi-dimensional pattern recognition | |
US4589140A (en) | Method of and apparatus for real-time high-speed inspection of objects for identifying or recognizing known and unknown portions thereof, including defects and the like | |
US6952491B2 (en) | Optical inspection apparatus for substrate defect detection | |
US5883982A (en) | Astigmatism measurement apparatus and method based on a focal plane separation measurement | |
US5422831A (en) | Acoustic and video imaging system for quality determination of pharmaceutical products | |
US5309374A (en) | Acoustic and video imaging system for quality determination of agricultural products | |
JPS6036940A (ja) | 光学検査システム | |
CN114662594B (zh) | 一种目标特征识别分析*** | |
JPH08189904A (ja) | 表面欠陥検出装置 | |
US5237512A (en) | Signal recognition and classification for identifying a fire | |
JP3752849B2 (ja) | パターン欠陥検査装置及びパターン欠陥検査方法 | |
EP0394348A1 (fr) | Procede et appareil d'evaluation optique d'objets inconnus | |
Tejedor et al. | Towards detection of pipeline integrity threats using a SmarT fiber-OPtic surveillance system: PIT-STOP project blind field test results | |
US20040258311A1 (en) | Method for generating geometric models for optical partial recognition | |
Niskanen et al. | Experiments with SOM based inspection of wood | |
Mkhoyan et al. | Adaptive real-time clustering method for dynamic visual tracking of very flexible wings | |
CN112508946A (zh) | 一种基于对抗神经网络的电缆隧道异常检测方法 | |
JPH04238207A (ja) | 欠陥検査装置 | |
FR2582831A1 (fr) | Procede d'identification d'une empreinte digitale, dispositif pour la mise en oeuvre de ce procede et utilisation de ce procede | |
CN116685827A (zh) | 表面粗糙度测量*** | |
CA2547304C (fr) | Methode de detection des defauts dans les grumes, y compris celle de la reconnaissance de formes multi-dimensionnelles | |
Hill et al. | Signal processing for automatic optical su rface inspection of steel strip | |
JPH03128412A (ja) | 自動選別装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
17P | Request for examination filed |
Effective date: 19900614 |
|
AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): CH DE FR GB IT LI SE |
|
A4 | Supplementary search report drawn up and despatched |
Effective date: 19920807 |
|
AK | Designated contracting states |
Kind code of ref document: A4 Designated state(s): CH DE FR GB IT LI SE |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE APPLICATION HAS BEEN WITHDRAWN |
|
18W | Application withdrawn |
Withdrawal date: 19930314 |