[go: up one dir, main page]

WO1996011386A1 - Procede de representation a base spectrale concernant des elements d'image disposes sur une ligne - Google Patents

Procede de representation a base spectrale concernant des elements d'image disposes sur une ligne Download PDF

Info

Publication number
WO1996011386A1
WO1996011386A1 PCT/SE1995/001068 SE9501068W WO9611386A1 WO 1996011386 A1 WO1996011386 A1 WO 1996011386A1 SE 9501068 W SE9501068 W SE 9501068W WO 9611386 A1 WO9611386 A1 WO 9611386A1
Authority
WO
WIPO (PCT)
Prior art keywords
rows
image
dimensional image
spectral
weighted
Prior art date
Application number
PCT/SE1995/001068
Other languages
English (en)
Inventor
Anders ÅSTRÖM
Erik ÅSTRAND
Original Assignee
Soliton Elektronik Ab
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
Application filed by Soliton Elektronik Ab filed Critical Soliton Elektronik Ab
Priority to EP95934904A priority Critical patent/EP0786074A1/fr
Priority to AU37119/95A priority patent/AU3711995A/en
Publication of WO1996011386A1 publication Critical patent/WO1996011386A1/fr

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/28Investigating the spectrum
    • G01J3/2803Investigating the spectrum using photoelectric array detector
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/28Investigating the spectrum
    • G01J3/2823Imaging spectrometer

Definitions

  • the present invention relates to a method of spectral-based representation of pixels on a line which is illuminated with light so as to produce a linear image, whereupon the linear image is transformed into a two-dimensional image in such a way that each column in the two-dimensional image can represent spectral divisions of the light which is reflected by corresponding points in the linear image, whereupon at least one total which can represent spectral characteristics for a pixel on the line is formed as a result of weighted pixels being totalized in columns.
  • the spectral-based representation of pixels is desirable when classifying objects on the basis of their colour.
  • the detected light can be divided into different spectral bands. If, for example, the object is red, the intensity for the divided light predominates principally in the red area.
  • the currently known method generally uses a conventional sensor, the entire two-dimensional image being read out and converted in an analogue-to-digital manner.
  • the object of the present invention is to provide a method for facilitating more rapid spectral-based representation of pixels on a line. This object is achieved in that a variable number of rows are read out from the two-dimensional image, and in that the image contents in the rows read-out are totalized in columns in an analogue manner.
  • the values of rows are weighted by using different exposure times for the rows in the two-dimensional image.
  • a weighted row value is approximated by analogue totalizing of the row to be weighted and rows lying in close proximity thereto.
  • FIG. 3 shows the manner in which different weighting can be attained by totalizing the frequency rows to be weighted and a number of rows adjacent thereto.
  • the reference sign 1 generally designates an optical system which comprises an objective 2, a gap 3, through which a linear image can be received, and a prism 4 by means of which the linear image can be transformed into a two-dimensional image.
  • the two-dimensional image is transferred to a two-dimensional optical sensor 5 which is preferably adapted to carry out given image- processing on the image detected by the sensor.
  • the sensor can be connected to any type of data system in which final evaluation is performed as necessary.
  • the procedure can be carried out by using an optical sensor which enables any rows in a two-dimensional image to be read out. Reading out can be of the type which is non ⁇ destructive but it can also be of the destructive type.
  • a two-dimensional optical sensor which has given inherent possibilities for signal processing
  • a sensor of the MAPP 2200 type produced by Integrated Vision Products AB.
  • This sensor consists of an array of a number of photodiodes and comprises a built-in processor which can carry out parallel processing of a row of pixels.
  • the sensor is constructed so that the rows to be processed can be selected in any order and read out without affecting the other data in the image. However, the reading out operation results in the data in the read-out rows being destroyed.
  • the transformed image can impinge on the two-dimensional array of photodiodes which forms the sensor element in the optical sensor (MAPP 2200).
  • the rows which are of especial interest are processed in the image-processing circuit. Since these rows in the sensor correspond to given frequencies or frequency ranges, the addressing in terms of rows entails given frequencies being removed for further processing. It will be appreciated that the actual frequencies are determined by the actual application. It is, therefore, important that the addressed rows can be varied. The total corresponding to a given spectral characteristic is accordingly obtained by analogue totalizing of the rows read out.
  • the frequencies in question in the recorded spectrum are preferably totalized in a weighted manner.
  • N s c (i) ⁇ j f(i,j) (l)
  • a number of totals is formed which correspond to a number of classes c.
  • the totals formed can, for example, be used directly for classifying the pixel by selecting the class c for which the above-mentioned total S c (i) is the greatest.
  • the frequency components can only occur once when this sensor is used for implementing the method, is that the data stored in the sensor are destroyed during subsequent image- processing.
  • the frequency components can be freely re-used in the totalizing process.
  • a further variant for calculating the totals S c (i) is to totalize in an analogue manner a number of rows in the vicinity of the frequency in question. In this way, it should be possible to to approximate 3 «f(i, x) according to the following
  • the preferred arrangement for implementing the method according to the invention cannot be used in the above manner when negative coefficients have to be processed in the totals S c (i) since negative exposure is not possible.
  • the signal-processing circuits are thus used in order to carry out this subtraction. For example, the positive coefficients are added first individually and the negative coefficients are added individually, whereupon necessary subtraction can be carried out in very different ways depending on which type of signal-processing circuits are used.
  • the method according to the invention can be regarded as a method of reducing pixel data by the formation of linear combinations.
  • a vector can be created for each pixel which has fewer dimensions than the original vector.
  • the magnitude in the case of this vector depends to a certain extent on the degree of precision required when the information is subsequently processed.
  • the information which is obtained when the method according to the invention has been applied is consequently a measure of the spectral characteristics of a large number of pixels; a number of reduced vectors to which further operations can be applied.
  • linear or non-linear classification can be used.
  • the classification can be brought about directly in the actual totalizing processes, only simple comparisons between the total values obtained having to be made.
  • the method according to the invention is suitable for rapid inspection and can, therefore, advantageously be used when material which is conveyed along a running track is inspected.
  • An application which is advantageous in this respect is when the camera is used for detecting defects in wood.
  • the method can be applied in particular for finding defects which cause the colour to differ from the surrounding wood, such as bruising, rotting, random marks, etc. Generally, differences in colour caused by these defects are too small to be detected using conventional colour camera technology whilst the speed requirement is too great for conventional spectrometry technology to be used at a reasonable cost.
  • a further suitable area of application is the inspection of paper.
  • the method according to the invention can advantageously be used for checking that paper running along a track is of the correct colour.

Landscapes

  • Physics & Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)
  • Image Processing (AREA)

Abstract

Un procédé de représentation à base spectrale concerne des pixels disposés sur une ligne qui est eclairée par une lumière pour produire une image linéaire, laquelle est transformée en une image bidimensionnelle de façon que chaque colonne de cette dernière représente des divisions spectrales de la lumière reflétée par des points correspondants de l'image linéaire. On obtient au moins un total qui peut représenter les caractéristiques spectrales d'un pixel de cette ligne en ajoutant des pixels pondérés propres à une colonne. Le procédé décrit se caractérise en ce qu'un nombre variable de rangées sont lues à partir de l'image bidimensionnelle, et en ce que la teneur d'image des rangées lues fait l'objet d'une totalisation analogique.
PCT/SE1995/001068 1994-10-10 1995-09-27 Procede de representation a base spectrale concernant des elements d'image disposes sur une ligne WO1996011386A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP95934904A EP0786074A1 (fr) 1994-10-10 1995-09-27 Procede de representation a base spectrale concernant des elements d'image disposes sur une ligne
AU37119/95A AU3711995A (en) 1994-10-10 1995-09-27 Method for a spectral based representation of image elements on a line

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
SE9403424A SE9403424L (sv) 1994-10-10 1994-10-10 Metod för en spektralbaserad representation av bildelement på en linje
SE9403424-6 1994-10-10

Publications (1)

Publication Number Publication Date
WO1996011386A1 true WO1996011386A1 (fr) 1996-04-18

Family

ID=20395536

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/SE1995/001068 WO1996011386A1 (fr) 1994-10-10 1995-09-27 Procede de representation a base spectrale concernant des elements d'image disposes sur une ligne

Country Status (4)

Country Link
EP (1) EP0786074A1 (fr)
AU (1) AU3711995A (fr)
SE (1) SE9403424L (fr)
WO (1) WO1996011386A1 (fr)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1991005360A1 (fr) * 1989-09-26 1991-04-18 The University Of Connecticut Spectrometre video a pixel
EP0490428A2 (fr) * 1990-12-07 1992-06-17 Philips Patentverwaltung GmbH Spectromètre à détection multicanale
US5173748A (en) * 1991-12-05 1992-12-22 Eastman Kodak Company Scanning multichannel spectrometry using a charge-coupled device (CCD) in time-delay integration (TDI) mode

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1991005360A1 (fr) * 1989-09-26 1991-04-18 The University Of Connecticut Spectrometre video a pixel
EP0490428A2 (fr) * 1990-12-07 1992-06-17 Philips Patentverwaltung GmbH Spectromètre à détection multicanale
US5173748A (en) * 1991-12-05 1992-12-22 Eastman Kodak Company Scanning multichannel spectrometry using a charge-coupled device (CCD) in time-delay integration (TDI) mode

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
R. FORCHHEIMER, P. INGELHAG, C. JANSSON, "MAPP+2200-A Second Generation Smart Optical Sensor", Presented at the SPIE Meeting in February 1992, San Jose, USA. *

Also Published As

Publication number Publication date
SE502197C2 (sv) 1995-09-11
AU3711995A (en) 1996-05-02
EP0786074A1 (fr) 1997-07-30
SE9403424D0 (sv) 1994-10-10
SE9403424L (sv) 1995-09-11

Similar Documents

Publication Publication Date Title
Jonckheere et al. Assessment of automatic gap fraction estimation of forests from digital hemispherical photography
RU2451340C2 (ru) Способ распознавания загрязнений и/или истираний краски в зоне переходов цветов на ценных документах и средства для осуществления этого способа
EP1484595B1 (fr) Transformation d'espaces de couleur pour identifier des objets d'intérêt dans des échantillons biologiques
GB2286670A (en) Detecting surface defects
Prasitmeeboon et al. Defect detection of particleboards by visual analysis and machine learning
WO1998021360A1 (fr) Procede et appareil de test microbiologique automatise
CN119723218A (zh) 异物识别方法、装置、设备及存储介质
Yoon et al. Hyperspectral imaging using RGB color for foodborne pathogen detection
Andreadis et al. Image pixel classification by chromaticity analysis
US6961476B2 (en) Autothresholding of noisy images
JPH08247954A (ja) 光学歪の測定方法
WO1996011386A1 (fr) Procede de representation a base spectrale concernant des elements d'image disposes sur une ligne
Grana et al. Associative morphological memories for endmember induction
US5912987A (en) Color detecting apparatus for inspecting a color of a printed matter
Wen et al. Brightness-invariant image segmentation for on-line fruit defect detection
JPH02236938A (ja) 画像復元方法並びに走査型電子顕微鏡及びパターン外観検査装置及び走査像検出装置
US20080283449A1 (en) Method of Analyzing and Sorting Eggs
Bharati Multivariate image analysis and regression for industrial process monitoring and product quality control
US20050017186A1 (en) Method and means for detecting internal larval infestation in granular material
Singh et al. Automated detection of plant leaf diseases using image processing techniques
JP4002387B2 (ja) 農産物の等級判定方法及び等級判定装置
JPH0760462B2 (ja) 画像領域分割装置
Xing et al. Combining multispectral reflectance and fluorescence imaging for identifying bruises and stem-end/calyx regions on Golden Delicious apples
JPH0676069A (ja) 表面欠陥検出装置
Szlavik et al. Visual inspection of metal objects by using cellular neural networks

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AM AT AU BB BG BR BY CA CH CN CZ DE DK EE ES FI GB GE HU IS JP KE KG KP KR KZ LK LR LT LU LV MD MG MK MN MW MX NO NZ PL PT RO RU SD SE SG SI SK TJ TM TT UA UG US UZ VN

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): KE MW SD SZ UG AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN ML MR NE SN TD TG

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
WR Later publication of a revised version of an international search report
WWE Wipo information: entry into national phase

Ref document number: 1995934904

Country of ref document: EP

REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

WWP Wipo information: published in national office

Ref document number: 1995934904

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: CA

WWW Wipo information: withdrawn in national office

Ref document number: 1995934904

Country of ref document: EP