Detecting, monitoring and controlling fouling of wetted surfaces

Abstract

A method and apparatus for repeatedly: acquiring images with more than three different frequency bands of a sampling object (210a, 210b, 210c) that is receiving a portion of water used in an industrial or municipal process; collecting visual data at a multitude of positions of the sampling object (210a, 210b, 210c) using the acquired images; performing spectral analysis of the collected visual data for the multitude of positions; and determining (404) a type of fouling of the sampling object (210a, 210b, 210c) based on the determined spectra.

Description

[0001] DETECTING, MONITORING AND CONTROLLING FOULING OF WETTED SURFACES

[0002] TECHNICAL FIELD

[0003] The present disclosure generally relates to a detecting, monitoring and controlling fouling of wetted surfaces. The disclosure relates particularly, though not exclusively, to detection and classification of fouling in water-intensive processes, based on multiband images of more than three frequency bands obtained in an on-line monitoring.

[0004] BACKGROUND

[0005] This section illustrates useful background information without admission of any technique described herein representative of the state of the art.

[0006] Various water processing plants use membranes to filter and treat water. Such membranes are prone to formation of deposits or fouling. The fouling adversely hinders flowing of water through the membranes so posing a problem in membrane filtration technologies. Membrane fouling comes in different categories including inorganic fouling or scaling, colloidal fouling, organic fouling, and biofouling. These tend to differ in cause, consequence, and cure: for example, biofouling involves forming of biological material development as a sticky layer on the membrane surface from deposition, growth and metabolism of bacteria cells or flocs on the membranes. Biofouling increases energy consumption through increased biofilm resistance and osmotic pressure and impairs quality of filtration due to increased solutes accumulation on the membrane surface. The biofouling can be controlled by suitable dosing of biocides. Inorganic fouling, on the other hand, may be controlled with anti-scaling agents, pH, or various process conditions suited to inhibit such fouling.

[0007] Kemira has developed a membrane fouling simulator, MFS, that a is a functional tool to predict the fouling potential in a non-destructive observation mode by utilising a bypass module. The MFS unit consists of transparent membrane cells connected to a pressure transmitter sensor. The fouling buildup is determined via monitoring pressure difference, and the extent and nature of fouling are recognised by visual observation. The visual observation may be carried out with an automated camera-based monitoring system that moves a camera over one or more test membranes and detects indicia of fouling from obtained images.

[0008] In camera based visual observation, identified shapes may be used to attempt determining the type or extent of fouling. Moreover, the colour distribution or spectra of identified shapes may be used to improve accuracy of foul detection and classification. However, while inexpensive, such camera-based detection faces physical limitations of the underlying phenomena: resolution, colour depth, and frequency bands: a digital camera forms pixel values on each of red, green, and blue channels by gathering light to respective sets of subpixels. More information can be obtained by alternating between different illuminants, e.g., between two or more known illuminants, such as white light and an illuminant that resides in or extends to wavelengths adjacent to the visible light wavelength band. The pixel granularity and colour granularity may also be partly linked: a visually perceivable shape differs by its colour from the background. At boundaries, the colour differences are divided into adjacent pixels such that some pixels may partially correspond to the colour of a shape and partially to the colour of the background.

[0009] Based on the detected fouling and fouling classification, suitable measures may be taken to inhibit the continuing of the detected fouling. To this end, an automated system may be formed including on-line monitoring and control, wherein the control attempts to inhibit the detected fouling based on measurement data obtained from the on-line monitoring.

[0010] It is desirable to improve existing fouling monitoring and classification as well as the fouling inhibiting by accompanying control, or to at least provide new alternatives to existing technology.

[0011] SUMMARY

[0012] The appended claims define the scope of protection. Any examples and technical descriptions of apparatuses, products and / or methods in the description and / or drawings not covered by the claims are presented not as embodiments of the invention but as background art or examples useful for understanding the invention.

[0013] According to a first example aspect there is provided an apparatus, comprising: a membrane fouling simulator comprising a sampling object configured to pass through a portion of water used in an industrial or municipal process; a multiband camera system configured to repeatedly acquire images with more than three different frequency bands of the sampling object to collect visual data at a multitude of positions of the sampling object; and at least one processor configured to repeatedly cause performing spectral analysis of the collected visual data for the multitude of positions; and determining a type of fouling of the sampling object based on the determined spectra. The portion of water may be received from an industrial process. Alternatively, the portion of water may be received from a municipal process. The municipal process may be an industrial process.

[0014] The determining of the type of the fouling may comprise determining whether type is one of a plurality of different fouling types. The fouling types may include biofouling. The fouling types may comprise organic fouling. The fouling types may comprise inorganic fouling.

[0015] The sampling object may comprise one membrane. The sampling object may comprise two or more membranes.

[0016] The collected visual data may define intensities of reflected colours in at least five different frequency bands.

[0017] The collected visual data may define intensities of reflected colours in a plurality of adjacent different frequency bands jointly spanning over a continuous spectral range.

[0018] The multiband camera system may comprise two or more cameras configured to form different colour channels of the collected visual data. The apparatus may comprise an image combiner configured to combine images taken by the two or more cameras.

[0019] The image combiner may be configured to align the images on the combining. The aligning may be based on position information of relative positions of individual cameras with which the images were taken. The aligning may be based on image analysis configured to align images by reducing difference between the images. The difference may involve brightness.

[0020] The two or more cameras may share one or more optical elements such as lenses or prisms. The two or more cameras may share a common housing. The two or more cameras may be arranged in a row.

[0021] The multiband camera system may be movable across the sampling object, optionally across entire membrane or membranes. The multiband camera system may be movable linearly. The multiband camera system may be movable along a longitudinal axis. The longitudinal axis may be parallel with the row in which the two or more cameras are arranged. The longitudinal axis may be movable, e.g., in a direction transverse to the longitudinal axes.

[0022] The spectral analysis may comprise comparing the collected visual data with one or more predefined indicia that are indicative of different types of fouling.

[0023] The indicia may comprise relative markers such as relative brightness of one or more frequency bands of the collected visual data in comparison to a brightness in some other frequency bands.

[0024] The indicia may comprise absolute markers such as absolute brightness of one or more frequency bands of the collected visual data.

[0025] The determining of the type of fouling may comprise detecting separate fouled portions indicated by the visual data. The determining of the type of fouling may comprise defining the type of fouling separately for each of the fouled portions or collectively based on two or more, optionally all, of the fouled portions.

[0026] The camera system may comprise a moveable cradle. The camera system may comprise one or more digital cameras attached to the cradle. The camera system may comprise an actuator coupled with the cradle and configured to move the cradle between a first position and a second position through a number of intermediate imaging positions.

[0027] The camera system may comprise an imaging alignment controller configured to cause the one or more digital cameras to capture images responsively to the cradle reaching any of the intermediate imaging positions.

[0028] The multiband camera system may comprise an adaptative optical filtering system configured to adapt optical filtering for the acquiring of the images by changing optical filters used. The adaptive optical filtering system may comprise a plurality of different optical filters for positioning into an optical path of the multiband camera system.

[0029] The optical filters may be configured to selectively pass light of given frequencies or frequency bands more than light of other frequencies or frequency bands.

[0030] The camera system may comprise an optical zoom mechanism configured to zoom in a camera view under control of the at least one processor so that images are obtained zooming in particular fouling specimens.

[0031] A quantity of the fouling may be determined with the type of fouling.

[0032] The at least one processor may be configured to perform a shape analysis of fouling objects visible on the sampling object. The at least one processor may be configured to perform the determining of the type of fouling determining a type of fouling of the sampling object based on the determined spectra and the shape analysis.

[0033] The at least one processor may be configured to repeat the causing of the performing of the spectral analysis and the determining of the type of fouling of the sampling object each time the multiband camera system performs the acquiring of the images. Alternatively, the at least one processor may skip the causing of the performing of the spectral analysis and the determining of the type of fouling of the sampling object sometimes the multiband camera system performs the acquiring of the images. The skipping may be performed for reducing processing. The skipping may be performed such that the performing of the spectral analysis and the determining of the type of fouling of the sampling object are performed with a fixed or varying interval. The interval may be at least 5 s, 10 s, 30 s, 1 minute, 2 minutes, 5 minutes, 10 minutes, or 30 minutes. The interval may be at most 30 s, 1 minute, 2 minutes, 5 minutes or 10 minutes, 1 hour, or 6 hours.

[0034] The apparatus may comprise a controller configured to control the industrial or municipal process according to the determining of the type of the fouling of the sampling object.

[0035] The controlling of the industrial or municipal process may comprise controlling addition of one or more chemical agents to the industrial or municipal process responsively to determining of a given type of fouling of the sampling object. The control of the addition of the one or more chemical agents may be configured to provide a feedback control of chemical addition. The control of the addition of the one or more chemical agents may be configured to provide a feedforward control of chemical addition.

[0036] The at least one processor may be configured to compute key variables are computed for the fouling, such as the fouling level and fouling rate for each type. The at least one processor may be configured to use the computed variables for causing at least one of: adjusting or optimising mechanical cleaning of membranes or filters in the process; adjusting or optimising mechanical cleaning chemical cleaning of membranes or filters in the process; calculating chemical dosages; optimising chemical programs; optimising adjustable parameters of chemical programs, such as recipes of chemicals; adjusting or optimising one or more process parameters such as pressure, flow rate, recovery, temperature, pH; optimising combinations of adjustable parameters of chemical programs, such as recipes of chemicals; or optimising dosing points of adjustable parameters of chemical programs, such as recipes of chemicals.

[0037] According to a second example aspect there is provided a method, comprising repeatedly: causing acquiring images with more than three different frequency bands of a sampling object that is receiving a portion of water used in an industrial or municipal process; collecting visual data at a multitude of positions of the sampling object using the acquired images; performing spectral analysis of the collected visual data for the multitude of positions; and determining a type of fouling of the sampling object based on the determined spectra.

[0038] The method may comprise controlling the industrial or municipal process according to the determining of the type of the fouling of the sampling object.

[0039] According to a third example aspect there is provided a computer program comprising computer executable program code which when executed by at least one processor causes an apparatus at least to perform the method of the second example aspect.

[0040] According to a fourth example aspect there is provided a computer program product comprising a non-transitory computer readable medium having the computer program of the third example aspect stored thereon.

[0041] According to a fifth example aspect there is provided an apparatus comprising means for performing the method of the second example aspect.

[0042] Any foregoing memory medium may comprise a digital data storage such as a data disc or diskette; optical storage; magnetic storage; holographic storage; opto-magnetic storage; phase-change memory; resistive random-access memory; magnetic random-access memory; solid-electrolyte memory; ferroelectric random-access memory; organic memory; or polymer memory. The memory medium may be formed into a device without other substantial functions than storing memory or it may be formed as part of a device with other functions, including but not limited to a memory of a computer; a chip set; and a sub assembly of an electronic device.

[0043] Different non-binding example aspects and embodiments have been illustrated in the foregoing. The embodiments in the foregoing are used merely to explain selected aspects or steps that may be utilised in different implementations. Some embodiments may be presented only with reference to certain example aspects. It should be appreciated that corresponding embodiments may apply to other example aspects as well.

[0044] BRIEF DESCRIPTION OF THE FIGURES

[0045] Some example embodiments will be described with reference to the accompanying figures, in which:

[0046] Fig. 1 schematically shows a system according to an example embodiment;

[0047] Fig. 2 shows a monitoring unit for monitoring scaling and fouling in a process, according to an example embodiment;

[0048] Fig. 3 shows a block diagram of an apparatus according to an example embodiment; and Figs. 4a to 4c show a flow chart of a process according to an example embodiment.

[0049] DETAILED DESCRIPTION

[0050] In the following description, like reference signs denote like elements or steps.

[0051] Fig. 1 schematically shows a system 100 according to an example embodiment. The system comprises a sample membrane 110, such as a reverse osmosis cell, through which some process water is filtered. A multiband camera system 120 is provided for collecting visual information from an upper surface 112 of the sample membrane. In an example embodiment, the multiband camera system captures images in at least four or five different colour channels. Most of the colour channels may reside in visible light range. Alternatively, a minority of the colour channels may reside in the visible light range. In an example embodiment, the number of colour channels is at least 10 or 50. The sample membrane 110 is provided with an input feed flow F, a first output C for a non-filtered concentrate flow, and a second output P for a filtered permeate flow. The second output P is present in this embodiment. In an example embodiment there is no such second output, as scaling and fouling deposits can be monitored and analysed also with impermeable or semipermeable sample membranes.

[0052] In an example embodiment, the system 100 further comprises a spacer layer applied on the upper surface 112 of the sample membrane. In an example embodiment, the visual data is then collected both from the spacer and the upper surface 112. In an example embodiment, this is accomplished by focusing a lens of the multiband camera system alternatingly on the upper surface 112 and the spacer. In an example embodiment, the multiband camera system a sufficient depth of field in the lens to capture sufficiently sharp images of both the upper surface 112 and the spacer.

[0053] In an example embodiment, the multiband camera system 120 collects information with illumination provided by one or more illuminators 130. In an example embodiment, the illuminators comprise one or more LED lamps, lasers, Xenon lights, UV lights, or halogen lights. In an example embodiment, the illumination is continuous. In an example embodiment, the illumination is intermittently flashing or strobe light, preferably temporally aligned with exposure times of the multiband camera system 120. In an example embodiment, the one or more illuminators 130 provide illumination of a desired range or ranges of wavelengths. In an example embodiment, white light illumination is provided for gaining information of the colour, brightness, shape and size of the fouling. In an example embodiment more than one illuminants operate simultaneously, optionally with different frequency ranges. In an example embodiment, the illuminators provide ultraviolet (LIV) light, visible light, infrared light, or any two or more of them.

[0054] The multiband camera system enhances type classification of the fouling. For example, some organic fouling absorbs UV-light making them appears as dark objects in respective frequency band. Some biofouling is fluorescent when excited by UV or some other suitable wavelengths. With a multiband camera system, the fluorescent emissions and their properties, such as the Stokes shift, may also or alternatively enable classifying the fouling.

[0055] The system 100 further comprises a data processing unit 140 that analyses the collected visual data. In an example embodiment, the data processing unit 140 classifies the quality of scaling and fouling on the receiving surface based on information obtained from the visual data and compares it with stored indicia in a digital library 150. Such a library may comprise a selection of comparison images or graphic representations of different scaling and fouling types, to which the visual data is compared. The comparison may employ image processing, such as image autocorrelation or artificial intelligence to determine closest match of histograms or spectra, or portions thereof, for example. In an example embodiment, an object detection API is used, such as api4ai, Chooch Al, Clarifai, Visua Al, Imagga, SentiSight.ai, or Hive Object Detection AP.

[0056] In an example embodiment, the indicia are obtained using known samples. This process may be performed by industrial installations by taking comparative laboratory samples and corresponding visual data. In an example embodiment, the suitable differentiating indicia are then determined through artificial intelligence classifier, image autocorrelation, or any other automatic method, or manually, e.g., by visually detecting deterministic portions of different spectra for each of different types of fouling.

[0057] The data processing unit 140 further computes a scaling and / or fouling indication or index which to be displayed on a display 160 or to be sent to an external circuitry for evaluation, control, or both. For example, the data processing unit 140 may send a control signal to control the process from which the process water is taken to the system sample membrane 110. In an example embodiment, the control signal is sent to a chemical dosing device 170 of the process. In an example embodiment, the control signal is sent to a pump for adjusting flow rate or pressure in the process. In an example embodiment, the control signal is sent to an actuator of a valve, heater, or other device that in part controls the process.

[0058] In an example embodiment, the process is desalination process of sea or brackish water, waste water, or circulated water. In an example embodiment, the process uses filtration units that are pressure driven membranes such as reverse osmosis membranes, nanofiltration membranes, ultrafiltration membranes, or microfiltration membranes.

[0059] Fig. 2 shows a monitoring unit 200 for monitoring scaling and fouling in a process, according to an example embodiment. It shows three cells or membranes 210a, 210b and 210c to be monitored and a multiband camera system 120 mounted on a cradle 130. An actuator is provided to move the multiband camera system 120 and one or more illuminators 130 across the 210a, 210b and 210c to collect visual across their surfaces. The multiband camera system 120, such as a pair or three digital CCD cameras equipped with a high- magnification lens and different image capturing frequency bands, can be moved longitudinally and laterally as shown by arrows A and B. In an example embodiment, the multiband camera system 120 is fixed to a given position over the cells 210a, 210b and 210c in a stable position, preferably with a relatively high resolution such as 50 to 200 million pixels. In an example embodiment, the multiband camera system 120 is fixed to a given position over the cells 210a, 210b, 210c with a tilting in one or two rotational directions (e.g., perpendicularly). In an example embodiment, the multiband camera system 120 is mounted on a linear guide. In an example embodiment, the actuator comprises a stepper motor to move the multiband camera system 120 between a first position and a second position through a number of intermediate positions for taking images at each position or some of these positions.

[0060] In an example embodiment, the multiband camera system 120 comprises an adaptative optical filtering system 122 configured to adapt optical filtering for the acquiring of the images by changing optical filters 124 used. In an example embodiment, the adaptative optical filtering system 122 comprises two or three different optical filters that are changed into use. In an example embodiment, the change of optical filter takes place by physically replacing the optical filter that is placed in an optical path of a camera or cameras used. In an example embodiment, the replacing of the optical filter takes place by moving the optical path to pass through a desired optical filter, e.g., by turning one or more prisms. In an example embodiment, different optical filters are simultaneously used by employing different frequency dependent total reflection of light to direct light from the measuring cell to a camera through a particular optical filter for a particular frequency band of light.

[0061] In an example embodiment, the multiband camera system 120 comprises an optical zoom mechanism 126 configured to in a camera view so that images are obtained zooming in particular fouling specimens.

[0062] In Fig. 2, the scaling and fouling is monitored with three separate measuring cells 210a, 210b and 210c. In an example embodiment, these measuring cells are connected in parallel to provide a larger sample of the same process step in a filtration plant. In an example embodiment, these measuring cells are connected in series to provide a larger surface area. Alternatively, the measuring cells may be connected to different flow streams and be used for showing the situation, e.g., in different steps of the filtration process. This is useful e.g. when studying effects of e.g. added anti-fouling chemicals or changed process parameters.

[0063] In one exemplary setup, a measuring cell is illuminated with white and LIV LED lights, the LIV wavelength being 395 nm, for example. A CCD camera and a unit for processing imaging data are also provided.

[0064] In an example embodiment, key variables are computed for the fouling, such as the fouling level and fouling rate for each type. In an example embodiment, the computed variables are used for monitoring and controlling the of the process or processes represented by respective measuring cell(s). In an example embodiment, the key variables are used to calculate chemical dosages and for optimising chemical programs, including adjustable parameters like recipes of chemicals, their combinations and dosing points. In an example embodiment, the key variables are used for adjusting or optimising one or more process parameters such as pressure, flow rate, recovery, temperature, pH.

[0065] The measuring cell may be provided with sample water through an electrically actuated ball control valve. In an example embodiment, the sample feed is controlled by a controller, such as a proportional-integral-derivative (PID) controller. In an example embodiment, the controller receives feedback from a liquid level sensor, such as an ultrasonic tank liquid level sensor. A pump such as a diaphragm pump is controlled by the controller to pump the sample through the monitoring cell and optionally through a back pressure valve out of the apparatus.

[0066] In an example embodiment, a programmable logic controller is used for monitoring and controlling operations, such as a Siemens S7-1200 PLC is used to control the operations of the analysing equipment. In an example embodiment, an industrial or general-purpose computer runs the computer program code that enables visual data processing.

[0067] Fig. 3 shows a block diagram of an apparatus 300 according to an example embodiment. In an example embodiment, the apparatus 300 is used to perform fouling classification based on visual data obtained from the multiband camera system 120. The apparatus 300 comprises a communication interface 310; a processor 320; a user interface 330; and a memory 340.

[0068] The communication interface 310 comprises in an embodiment a wired and / or wireless communication circuitry, such as Ethernet; Wireless LAN; Bluetooth; GSM; CDMA; WCDMA; LTE; and / or 5G circuitry. The communication interface can be integrated in the apparatus 300 or provided as a part of an adapter, card or the like, that is attachable to the apparatus 300. The communication interface 310 may support one or more different communication technologies. The apparatus 300 may also or alternatively comprise more than one of the communication interfaces 310.

[0069] In this document, a processor may refer to a central processing unit (CPU); a microprocessor; a digital signal processor (DSP); a graphics processing unit; an application specific integrated circuit (ASIC); a field programmable gate array; a microcontroller; or a combination of such elements.

[0070] The user interface may comprise a circuitry for receiving input from a user of the apparatus 300, e.g., via a keyboard; graphical user interface shown on the display of the apparatus 300; speech recognition circuitry; or an accessory device; such as a headset; and for providing output to the user via, e.g., a graphical user interface or a loudspeaker.

[0071] The memory 340 comprises a work memory 342 and a persistent memory 344 configured to store computer program code 346 and data 348. The memory 340 may comprise any one or more of: a read-only memory (ROM); a programmable read-only memory (PROM); an erasable programmable read-only memory (EPROM); a random-access memory (RAM); a flash memory; a data disk; an optical storage; a magnetic storage; a smart card; a solid- state drive (SSD); or the like. The apparatus 300 may comprise a plurality of the memories 340. The memory 340 may be constructed as a part of the apparatus 300 or as an attachment to be inserted into a slot; port; or the like of the apparatus 300 by a user or by another person or by a robot. The memory 340 may serve the sole purpose of storing data or be constructed as a part of an apparatus 300 serving other purposes, such as processing data.

[0072] A skilled person appreciates that in addition to the elements shown in Fig. 3, the apparatus 300 may comprise other elements, such as microphones; displays; as well as additional circuitry such as input / output (I / O) circuitry; memory chips; application-specific integrated circuits (ASIC); processing circuitry for specific purposes such as source coding / decoding circuitry; channel coding / decoding circuitry; ciphering / deciphering circuitry; and the like. Additionally, the apparatus 300 may comprise a disposable or rechargeable battery (not shown) for powering the apparatus 300 if external power supply is not available.

[0073] In an example embodiment, the apparatus 300 illustrated in Fig. 3 further illustrates some elements of a controller suited for controlling a process. There may be a plurality of different apparatuses that can be illustrated by the same drawing. In an example embodiment, one apparatus may function in a two or more roles, e.g., based on suitable programming to implement such different functions.

[0074] Some example embodiments regarding visual information processing are next briefly discussed. In an example embodiment, a Bayesian - Laplace probabilistic classification approach is used for fouling classification, which is advantageously a robust and well-suited method to discriminate different species of deposits from each other. In an example embodiment, all objects are classified to one specific object or particle class, like filaments, deposits of crystals, scales and other fouling objects. The classification may also rely on a hypercube approach, which means that a particle is classified to a particle class when particle’s every property remains between the discrete minimum and maximum limits specified for the class.

[0075] In an example embodiment, the classification comprises a scheme including phases 1 - 3 described in the following.

[0076] 1 ) Image filtering

[0077] Image filtering is utilised to remove noise, to fade out an unequal background, to highlight the focused objects, and to compute e.g. local greyscale gradient values and their direction. A filtered image may then be equalised e.g. by multiresolution analysis, e.g. using a Gaussian multiresolution pyramid. A Laplacian image (which is the second derivative of image greyscales) may then be computed from an equalised image to highlight the regions of the greatest greyscale variance.

[0078] 2) Image segmentation

[0079] The purpose of an image segmentation step is to recognise focused objects in an image and to compute the projective areas and outlines of the objects, and to recognise different types of objects in such image.

[0080] Dark regions are recognised by applying a greyscale percentile threshold to a cumulative greyscale histogram of an equalised image. The background of an image may be computed as the mean image of the previous 10 images. Thus, structural components of the area to be monitored, like spacers, may be digitally masked at an early stage from the segmentation analysis of the image.

[0081] Deposits, i.e. stagnant objects that are slowly building up, are identified from the image using the above-mentioned greyscale percentile threshold. The total area of the deposited objects per total image area x100% may be used as an indicator of a current fouling value.

[0082] Focus discrimination on a Laplacian image may be used to validate objects. Objects which projected area has more focused pixels relative to the total area than a user-specified focus ratio (e.g. 7 %), are recognised as valid. Regions of high greyscale variance may be highlighted by combining Laplacian, gradient and high pass filtered images. A binary image of the objects is obtained by applying to the combined image a user-specified contrast threshold and by superimposing on the image the dark regions.

[0083] 3) object morphology

[0084] A binary image of an object may be processed with morphological operations. As the projective area of each object is imaged by the camera, the object diameter d is defined based on the object’s projective area A as:

[0085] The morphology of objects may further be studied by defining their shape properties, including the aspect ratio, roundness, and coarseness.

[0086] When an object is recognised as an elongated object, an analysis may be carried out to obtain the length and width of the object. An analysis algorithm may be used, where the object length may be computed as the length of the outline (perimeter) divided by two. The width computation may be based on outline vectors consisting of the x, y -coordinates and the greyscale gradient direction value [-n, n] of each outline pixel. A matching point at the opposite side of the image outline is searched by comparing the direction values of the opposite outline pixels and of a line drawn between the matching pixels. The distance between the opposite pixels corresponds to the local width of an object, the overall width of which may then be computed as the mean of all local widths.

[0087] The principal axes and aspect ratio of deposits may be computed from the object by using principal component analysis (PCA) algorithm. The algorithm returns the major and minor axes of the object and their orientation angle. The aspect ratio may be computed as simply the ratio between the major and minor axes of the object.

[0088] Roundness describes how close to a circle an object is. A perfect circle has a roundness of 100 %. The roundness percentage decreases with an increasing complexity of the particle shape. Roundness R is computed as: where r is the object radius and are^g distances from outline pixels

[0089] (Xj, yi) to the centre point of the object (xc, yc). N= perimeter length.

[0090] The normalisation is obtained by dividing the standard deviation of radii with the object radius.

[0091] The coarseness of an object may be computed as the sum of discrete curvatures along the outline of the object divided by the length of the outline. Curvature values may be computed as a difference between the greyscale gradient direction angles of neighbouring outline pixels. Only rapid turns in the curvature are counted in the coarseness computation. The coarseness value may be normalised with the perimeter value of a circle having the same diameter as the maximum distance across the object. Kurtosis can be calculated by using 4th momentum of grey scale intensity. This can be used for can be used for classification of fouling type.

[0092] All detected objects in the receiving surfaces are classified to a one specific fouling type (e.g. biofouling, organic fouling and inorganic fouling or their combination(s)) according to predetermined classification criteria. Classification criteria may also include colours detectable from deposits by using white, ultraviolet or fluorescence excitation light, alone or in combination. Notably, colours in the present document are not restricted to visible light sub-bands but include also sub-bands beyond the visible light spectrum as enabled by the multi-band camera system that may extend is image capture range beyond the visible light range.

[0093] The texture of an object is important for cognitive recognition. Texture analysis may be done by modelling the object texture by studying the brightness (i.e. greyscale intensity) profile from the object centre point to its outline. The mean brightness values are computed at the particle centre, at the particle outline and at the full particle area. Also, the standard deviation of particle brightness values is computed. The mean brightness values may be utilised to discriminate particles to bright and dark classes and to classify bright and thin objects.

[0094] Application areas

[0095] In an example embodiment, the fouling classification is applied in the paper industry and its water streams. Other examples include, without restriction, oil, mining or water treatment processes, in particular desalination processes, membrane processes, cooling water treatment, and water reuse. For example, in water intensive industries, the subjects for monitoring efforts are organic, inorganic and biofouling, and combinations thereof.

[0096] Some example embodiments are used both for monitor and control of the water-intensive processes involved, and thus to control the addition rate of one or more process chemicals. In an example embodiment, the controlling is carried out manually. In an example embodiment, the controlling is semi-automatic. In an example embodiment, the controlling is automatic.

[0097] In the method, visual data may be collected at a multitude of positions across a receiving surface, and the visual data is analysed and classified to determine the quality and type of deposition attached to the receiving surface. In the method it is possible to recognise and classify different fouling types. Fouling type may be inorganic, organic or biofouling. The used deposition classification schemes may be based on object size, shape, texture and colour. The method enables measuring the properties of several fouling deposits. It discloses how to identify and classify several different fouling deposits and enables detecting multiple foulants and classification of foulants attached to the same receiving surface. In the method, actual deposits of all kinds may be monitored, classified and reported. These deposits may include organic, inorganic, and / or biofouling.

[0098] Figs. 4a to 4c show a flow chart according to an example embodiment, illustrating a process comprising various possible steps including some optional steps while also further steps can be included and / or some of the steps can be performed more than once:

[0099] 401. causing acquiring images with more than three different frequency bands of a sampling object that is receiving a portion of water used in an industrial or municipal process;

[0100] 402. collecting visual data at a multitude of positions of the sampling object using the acquired images;

[0101] 403. performing spectral analysis of the collected visual data for the multitude of positions; and 404. determining a type of fouling of the sampling object based on the determined spectra.

[0102] In an example embodiment, the process further comprises one or more of following:

[0103] 405. in determining of the type of the fouling, determining whether type is one of a plurality of different fouling types including biofouling;

[0104] 406. the collected visual data defining intensities of reflected colours in at least five different frequency bands;

[0105] 407. the collected visual data defining intensities of reflected colours in a plurality of adjacent different frequency bands jointly spanning over a continuous spectral range;

[0106] 408. forming a multiband camera system using two or more cameras configured to form different colour channels of the collected visual data;

[0107] 409. combining images taken by the two or more cameras;

[0108] 410. aligning the images on the combining;

[0109] 411. moving the multiband camera across the sampling object, optionally across entire membrane or membranes;

[0110] 412. in the spectral analysis, comparing the collected visual data with one or more predefined indicia that are indicative of different types of fouling;

[0111] 413. in the determining of the type of fouling, comprise detecting separate fouled portions indicated by the visual data;

[0112] 414. in the determining of the type of fouling, defining the type of fouling separately for each of the fouled portions or collectively based on two or more, optionally all, of the fouled portions;

[0113] 415. using an adaptative optical filtering system to adapt optical filtering for the acquiring of the images by changing optical filters used;

[0114] 416. using an optical zoom mechanism configured to zoom in a camera view under control of the at least one processor so that images are obtained zooming in particular fouling specimens;

[0115] 417. determining a quantity of the fouling with the type of fouling;

[0116] 418. performing a shape analysis of fouling objects visible on the sampling object;

[0117] 419. controlling the industrial or municipal process according to the determining of the type of the fouling of the sampling object;

[0118] 420. in the controlling of the industrial or municipal process, controlling addition of one or more chemical agents to the industrial or municipal process responsively to determining of a given type of fouling of the sampling object;

[0119] 421. in the controlling of the addition of the one or more chemical agents, providing a feedback control of the chemical addition;

[0120] 422. in the controlling of the addition of the one or more chemical agents, provide a feedforward control of the chemical addition;

[0121] 423. computing key variables for the fouling, such as the fouling level and fouling rate for each type;

[0122] 424. using the computed variables for causing at least one of: adjusting or optimising mechanical cleaning of membranes or filters in the process such as running a brushing or water jet cleaning; adjusting or optimising mechanical cleaning chemical cleaning of membranes or filters in the process, such as applying chemical cleaning agent on the membranes or filters; adjusting or optimising one or more process parameters such as pressure, flow rate, recovery, temperature, pH; calculating chemical dosages; optimising chemical programs; optimising adjustable parameters of chemical programs, such as recipes of chemicals; optimising combinations of adjustable parameters of chemical programs, such as recipes of chemicals; or optimising dosing points of adjustable parameters of chemical programs, such as recipes of chemicals.

[0123] Any of the afore described methods, method steps, or combinations thereof, may be controlled or performed using hardware; software; firmware; or any combination thereof. The software and / or hardware may be local; distributed; centralised; or any combination thereof. Moreover, any form of computing, including computational intelligence, may be used for controlling or performing any of the afore described methods, method steps, or combinations thereof. Computational intelligence may refer to, for example, any of artificial intelligence; neural networks; fuzzy logics; machine learning; genetic algorithms; evolutionary computation; or any combination thereof.

[0124] Various embodiments have been presented. It should be appreciated that in this document, words comprise; include; and contain are each used as open-ended expressions with no intended exclusivity.

[0125] The foregoing description has provided by way of non-limiting examples of particular implementations and embodiments a full and informative description of the best mode presently contemplated by the inventors for carrying out the invention. It is however clear to a person skilled in the art that the invention is not restricted to details of the embodiments presented in the foregoing, but that it can be implemented in other embodiments using equivalent means or in different combinations of embodiments without deviating from the characteristics of the invention.

[0126] Furthermore, some of the features of the afore-disclosed example embodiments may be used to advantage without the corresponding use of other features. As such, the foregoing description shall be considered as merely illustrative of the principles of the present invention, and not in limitation thereof. Hence, the scope of the invention is only restricted by the appended patent claims.

Claims

CLAIMS1. An apparatus, comprising: a membrane fouling simulator (200) comprising a sampling object (210a, 21 Ob, 210c) configured to pass through a portion of water used in an industrial or municipal process; a multiband camera system (210) configured to repeatedly acquire (401 ) images with more than three different frequency bands of the sampling object to collect visual data at a multitude of positions of the sampling object; and at least one processor (320) configured to repeatedly cause performing (403) spectral analysis of the collected visual data for the multitude of positions; and determining (404) a type of fouling of the sampling object (210a, 210b, 210c) based on the determined spectra.

2. The apparatus of claim 1 , wherein the determining (404) of the type of the fouling comprises determining (405) whether type is one of a plurality of different fouling types including biofouling.

3. The apparatus of claim 1 , wherein the collected visual data defines intensities of reflected colours in at least five different frequency bands.

4. The apparatus of claim 1 or 2, wherein the collected visual data defines intensities of reflected colours in a plurality of adjacent different frequency bands jointly spanning over a continuous spectral range.

5. The apparatus of any one of preceding claims, wherein the multiband camera system (210) comprises two or more cameras configured to form different colour channels of the collected visual data; and the apparatus further comprises an image combiner configured to combine images taken by the two or more cameras.

6. The apparatus of any one of preceding claims, wherein the spectral analysis comprises comparing the collected visual data with one or more predefined indicia that are indicative of different types of fouling.

7. The apparatus of claim 6, wherein the indicia may comprise relative markers such as relative brightness of one or more frequency bands of the collected visual data in comparison to a brightness in some other frequency bands.

8. The apparatus of any one of preceding claims, wherein the determining (404) of the type of fouling comprisesdetecting (413) separate fouled portions indicated by the visual data; and defining (414) the type of fouling separately for each of the fouled portions or collectively based on two or more, optionally all, of the fouled portions.

9. The apparatus of any one of preceding claims, wherein the multiband camera system (210) comprises an adaptative optical filtering system configured to adapt (415) optical filtering for the acquiring of the images by changing optical filters used.

10. The apparatus of any one of preceding claims, further comprising determining (417) with the type of fouling also a quantity of the fouling.11 . The apparatus of any one of preceding claims, the at least one processor (320) further being further configured to perform shape analysis of fouling objects visible on the sampling object (210a, 210b, 210c); and the at least one processor (320) further being further configured to perform the determining (404, 418) of the type of fouling determining a type of fouling of the sampling object (210a, 210b, 210c) based on the determined spectra and the shape analysis.

12. The apparatus of any one of preceding claims, further comprising a controller (300) configured to control the industrial or municipal process according to the determining (404) of the type of the fouling of the sampling object (210a, 210b, 210c).

13. The apparatus of any one of preceding claims, the at least one processor (320) being further configured to compute key variables are computed for the fouling, such as the fouling level and fouling rate for each type; and the at least one processor (320) being further configured to use the computed variables for causing (424) at least one of: adjusting or optimising mechanical cleaning of membranes or filters in the process; adjusting or optimising mechanical cleaning or chemical cleaning of membranes or filters in the process; calculating chemical dosages; optimising chemical programs; optimising adjustable parameters of chemical programs, such as recipes of chemicals; adjusting or optimising one or more process parameters such as pressure, flow rate, recovery, temperature, pH; optimising combinations of adjustable parameters of chemical programs, such as recipes of chemicals; oroptimising dosing points of adjustable parameters of chemical programs, such as recipes of chemicals.

14. A method, comprising repeatedly: acquiring images with more than three different frequency bands of a sampling object (210a, 210b, 210c) that is receiving a portion of water used in an industrial or municipal process; collecting visual data at a multitude of positions of the sampling object (210a, 210b, 210c) using the acquired images; performing (403) spectral analysis of the collected visual data for the multitude of positions; and determining (404) a type of fouling of the sampling object (210a, 210b, 210c) based on the determined spectra.

15. The method of claim 14, further comprising controlling the industrial or municipal process according to the determining (404) of the type of the fouling of the sampling object (210a, 210b, 210c).

Images