Estimating biofilm biomass on objects in an aquatic environment

Abstract

System and method for estimating biofilm biomass originating from an aquatic environment on a coating of an object. The method comprises obtaining one or more digital images of a soiled portion of a coating on an object and determining from each of the one or more images a respective reflectance value for the portion of the coating. The method comprises determining from the one or more reflectance values a value of a spectral index representative of the biomass and calculating a biomass pigment surface area density from the spectral index SI and one or more calibration values determined for a reference coating. The method comprises compensating the calculated biomass pigment surface area density for the reflectance of the coating of the object by applying a compensation associated with a difference in reflectance of the coating of the object relative to the reference coating.

Description

Technical Field

[0001] This invention relates to a method and system for estimating the biomass of biofilm fouling on objects in an aquatic environment. Background Technology

[0002] Fouling is a common problem for structures that are permanently or intermittently immersed in aquatic environments. Such structures include the hulls and ballast tanks of ships, vessels, and yachts; fixed and floating structures used for offshore oil and gas exploration, production, and storage; offshore structures used for wind and wave power generation; and conduits and pipelines.

[0003] Problems related to fouling include increased fuel costs for vessels due to increased frictional resistance. On static structures, such as drilling rigs, it can alter water flow around support legs, risking unpredictable and increased stress. It can also conceal defects and cracks during inspections. Fouling can further reduce the cross-sectional area of ​​piping works, such as cooling water or ballast tank inlets, resulting in reduced flow velocities. Fouling control coatings can be used to reduce fouling growth, but such coatings cannot completely prevent bioadhesion or biofilm formation.

[0004] Fouling can be broadly classified into two categories: macro-fouling (such as barnacles) and micro-fouling (biofilm / slime). Quantitative methods for measuring the amount of fouling (i.e., macro-fouling and micro-fouling) present on objects in aquatic environments can help estimate additional drag and / or determine the need for cleaning and / or recoating of the object.

[0005] Biofilms responsible for microfouling typically consist of pigmented microalgae and / or bacteria. These often contain photosynthetic pigments such as chlorophyll a. Chlorophyll a absorbs light very strongly at approximately 673 nm, and biofilms tend to be almost transparent in the near-infrared wavelength range. Therefore, it is possible to estimate the amount of chlorophyll in a typical fouling biofilm sample by measuring how much red light it reflects and often comparing that to a baseline measurement of near-infrared reflectance (depending on which biomass index is used).

[0006] Although quantitative imaging of biofilms has been previously performed, biofilms are often treated almost as homogeneous materials with characteristic absorption properties. These models typically fail to provide satisfactory results and / or are not applicable to many different subjects. Therefore, there is a need to improve the quantitative determination of biofilm fouling. Summary of the Invention

[0007] The object of this invention is to provide a method and system for estimating the biomass of biofilms originating from an aquatic environment on a coating of an object, such as an object that is permanently or intermittently immersed in an aquatic environment. This method and system are relatively accurate, applicable to a variety of objects, and easy to implement.

[0008] One aspect of the invention provides a method for estimating biofilm biomass derived from an aquatic environment on a coating, such as a method for estimating biofilm biomass derived from an aquatic environment on an object permanently or intermittently immersed in an aquatic environment. Such objects may be, for example, the hull or ballast tank of a ship, vessel, or yacht; fixed and floating structures for offshore oil or gas exploration, production, and storage; offshore structures for wind or wave power generation; or conduits or pipelines. The aquatic environment may be, for example, a marine, freshwater, or brackish water environment.

[0009] The method includes step a): obtaining one or more digital images of a contaminated portion of a coating on an object. Each of the one or more digital images can be obtained within a spectral band. A spectral band can, for example, have a spectral width in the range of 0-100 nm or greater. Here, a spectral width of 0 nm represents a single wavelength. Therefore, a single wavelength in this document is also considered a spectral band. A spectral band can, for example, have a spectral width in the range of 0-100 nm FWHM (full width at half maximum) or greater. Different digital images can be obtained within the same or different spectral bands. Each digital image may include one or more pixels.

[0010] The method includes step b): determining the corresponding reflectance value of the portion of the coating from each of one or more images.

[0011] The method includes step c): determining the value of a spectral index representing biomass based on one or more reflectance values.

[0012] The method includes step d): calculating the biomass pigment surface area density based on the spectral index SI and one or more calibration values ​​determined for a reference coating. SI The reference coating may differ from the object's coating, for example, by having a different color. While the calculated biomass pigment surface area density can be a measure of biofilm biomass, it has been found that the determined density value is affected by the reflectivity of the object's underlying coating. Therefore, this determined density is not a reliable indicator of biofilm biomass, as it may overestimate or underestimate biomass depending on the reflectivity of the object's coating.

[0013] Therefore, the method includes step e): compensating the biomass pigment surface area density calculated for the reflectance of the coating on the object by applying a compensation associated with the difference in reflectance of the coating on the object relative to a reference coating.SI Therefore, the estimated biomass density on the current coating surface is obtained. current Therefore, the reflectivity of the underlying coating of the object is compensated, allowing for more accurate determination of biofilm biomass. Furthermore, according to this method, it is not necessary to calibrate the biomass pigment surface area density (biomass) for each coating. SI Along with spectral indices, a reference coating can be calibrated, and then the difference between the actual coating on the object and the reference coating can be compensated. This method can be implemented using a computer.

[0014] Optionally, compensation is based on a comparative reflectance value of a portion of the coating representing the biofilm biomass. Therefore, the reflectance values ​​of the coating beneath the biomass in one or more images can be considered by comparing reflectance values. The comparative reflectance value can be a reflectance value determined for a portion of the coating from which all biomass has been removed; a reflectance value determined from a reference object; or a reflectance value stored in a database. The comparative reflectance value can be the minimum of the reflectance values ​​for the coating beneath the biofilm biomass in one or more spectral bands from which one or more images were obtained.

[0015] Therefore, in step d), the biomass pigment surface area density (biomass) can be determined. SI This is as if the current coating were a reference coating with a reference reflectance spectrum. Then, in step e), the difference in reflectance between the current coating and the reference coating can be compensated for by adjusting the biomass pigment surface area density determined for the reference coating. SI To estimate the biomass density on the current coating surface. current Optionally, the reflectivity of a reference coating is determined at a predetermined wavelength, and the reflectivity of the current coating is determined at the same predetermined wavelength, for determining the difference in reflectivity between the current coating and the reference coating.

[0016] Optionally, biomass is estimated based on spectral indices and one or more calibration values. SI It can be achieved through the error factor EF current and baseline error BE current Adjustments are made for both, which are functions of the reflectance spectrum of the current coating on which biomass exists. Then, the estimated biomass on the current coating is calculated. current This can be determined from the following equation:

[0017]

[0018] Optionally, the spectral band is 100 nm wide, FWHM or smaller, such as 50 nm wide, FWHM or smaller, such as 30 nm wide, FWHM or smaller, such as 5 nm FWHM or smaller. The spectral band can be at least 0.1 nm, for example, at least 1 nm. Therefore, the spectral band can be in the range of 0.1 to 100, 0.1 to 50, 0.1 to 30, or 0.1 to 5, such as from 1 to 100, from 1 to 50, or from 1 to 30, or from 1 to 5.

[0019] Optionally, one of the spectral bands is selected to include the absorption wavelength of chlorophyll, such as chlorophyll a. When obtaining a single digital image of a contaminated portion of a coating on an object, the spectral band can be selected to include the absorption wavelength of chlorophyll. When obtaining two or more images of a contaminated portion of a coating on an object, at least one of those images can be obtained in a spectral band that includes the absorption wavelength of chlorophyll. When obtaining two or more digital images of a contaminated portion of a coating on an object, at least one image can be obtained in a spectral band selected to exclude the absorption wavelength of chlorophyll. Therefore, the presence of biomass can be determined by the absorption of light by chlorophyll.

[0020] In this embodiment, the spectral band is located in the range of 350 nm to 900 nm, for example, in the range of 400 nm to 850 nm, meaning the spectral band includes one or more wavelengths falling within the aforementioned range. Optionally, at least one of the spectral bands near 433, 460, 496, 555, 584, 601, 673, or 800 nm is selected, i.e., such that the spectral band is or includes any specified wavelength. At least one of the spectral bands can be selected within the ranges of 390-480 nm, 410-510 nm, 440-550 nm, 500-610 nm, 525-645 nm, 540-660 nm, 605-740 nm, or 720-900 nm.

[0021] Optionally, the spectral index determined in step c) includes the ratio of two reflectance values. The ratio of reflectance values ​​allows for the normalization of the reflectance values, for example, relative to the reflectance of the object coating. Optionally, the spectral index determined in step c) includes the sum and / or difference of two reflectance values. One of the reflectance values ​​can be used for a spectral band that includes the absorption wavelengths of chlorophyll. The other of the reflectance values ​​can be used for a spectral band that does not include the absorption wavelengths of chlorophyll. Therefore, the relative effect of chlorophyll on light absorption can be explained.

[0022] Optionally, the spectral index SI determined in step c) is one of the following:

[0023] - The normalized red reflectance of the coating is defined as follows:

[0024] or

[0025] - Normalized Difference Vegetation Index (NDVI), which is defined as

[0026]

[0027] However, it should be understood that other spectral indices, such as (but not limited to) the benthic microalgae index, can be used, which is defined as Ocean Color Index, which is defined as Or the diatom index, which is defined as In this paper, R433 is a reflectance value determined in the range of 390-480 nm (e.g., at 433 nm); R460 is a reflectance value determined in the range of 410-510 nm (e.g., at 460 nm); R496 is a reflectance value determined in the range of 440-550 nm (e.g., at 496 nm); and R555 is a reflectance value determined in the range of 500-610 nm (e.g., at 555 nm). R584 is a reflectance value determined in the range of 525-645 nm (e.g., at 584 nm); R601 is a reflectance value determined in the range of 540-660 nm (e.g., at 601 nm); R673 is a reflectance value determined in the range of 605-740 nm (e.g., at 673 nm); R800 is a reflectance value determined in the range of 720-900 nm (e.g., at 800 nm); R673 clean This refers to the reflectance value in the 605-740 nm range (e.g., at 673 nm) at the coating on an object without biomass, such as the reflectance value determined at a clean portion of the object, a reference reflectance value with a coating identical or similar to the object, or a stored reference reflectance value. Spectral indices can be used to estimate the biomass pigment surface area density in step d). It has been found that the spectral index of the coating's normalized red reflectance and the normalized differential vegetation index provide good estimates of the biomass pigment surface area density.

[0028] As can be seen from the above, the normalized red reflectance of the coating can be determined based on a single digital image of the contaminated portion of the coating on the object within the spectral band, as well as the data used for R673. clean The stored values ​​are used to determine the spectral index. In other instances, the spectral index is determined from two or more digital images of a contaminated portion of the coating on an object in mutually different spectral bands. The two or more images can be separate layers of hyperspectral images.

[0029] The calculation of biomass pigment surface area density in step d) is based on one or more calibration values. These calibration values ​​can be determined for a reference coating having a predetermined reflectance. According to the invention, the reference coating does not need to be the current coating on the object. Step e) is provided to compensate the biomass pigment surface area density calculated based on the current coating for the reflectance of the object's coating by applying a compensation associated with the reflectance of the object's coating relative to the reference coating.

[0030] Optionally, in step d), the pigment surface area density is determined according to the following equation.

[0031]

[0032] Where SI is the spectral index value, and a, b, and c are calibration values. Calibration values ​​a, b, and c can be determined for a reference coating with a predetermined reflectivity.

[0033] In the example, calibration values ​​a, b, and c have been determined for a reference coating with gray color and a reflectivity value of 24%. In this case, the following calibration values ​​have been found.

[0034] -If the spectral index is the coating's normalized red reflectance...

[0035] a = -1.921, b = 0.093, c = 2.477; and

[0036] -If the spectral index is a normalized differential vegetation

[0037] a=0.618, b=0.240, c=-0.321.

[0038] Because these calibration values ​​result in the estimated biomass pigment surface area density in step d), SI If the current coating is a reference coating, then in step e), the estimated biomass pigment surface area density (biomass) is applied. SI The difference in reflectance of the current coating relative to the reference coating can be compensated by applying a compensation associated with the reflectance of the current coating relative to the reference coating. As indicated above, the estimated biomass on the current coating... current This can be determined from equation EQ1. Therefore, when combining equations EQ1 and EQ2, the biomass pigment surface area density (biomass) for the current coating is... current This can be determined from the following equation:

[0039]

[0040] It will be understood that steps d) and e) can be performed as individual steps or as a combination of steps. For example, when steps d) and e) are performed individually, equations EQ1 and EQ2 can be used. For example, when steps d) and e) are performed as a combination of steps, equation EQ3 can be used.

[0041] The following exemplary error factor EF has been found. current and baseline error BE current value:

[0042] -If the spectral index is the coating's normalized red reflectance...

[0043] EF current =0.26ln(R) current673 -1.00; BE current =0.0862ln(R) current 673 +0.0164;

[0044] -If the spectral index is a normalized differential vegetation

[0045] EF current =0.24ln(R) current673 -0.83; BE current =0.025ln(R) current 673 +0.1025;

[0046] Among them, R current673 This is the reflectance value of the current coating at 673nm.

[0047] Using these calibration curves, the error factor EF for a specific coating on an object can be determined. current and baseline error BE current Therefore, calibration values ​​a, b, and c determined for the reference coating, along with an error factor EF to compensate for the difference in reflectivity of the current coating relative to the reference coating, are used. current and baseline error BE current The value is used to determine the biomass pigment surface area density for the current coating. current The estimate is possible.

[0048] Optionally, one or more digital images may be obtained using a submarine (such as an unmanned submarine). The submarine may include digital cameras, such as hyperspectral cameras. Alternatively or additionally, one or more digital images may be obtained by divers, hull crawlers, imaging units that can descend along the object, photo booths (in dry dock), etc.

[0049] According to another aspect of the invention, a system is provided for estimating biomass of aquatic-derived biofilm on a coating on an object, for example, permanently or intermittently immersed in an aquatic environment. The system includes a processor. The processor is configured to acquire one or more digital images of contaminated portions of the coating on the object. The processor is configured to determine a corresponding reflectance value for the portion of the coating from each of the one or more images. The processor is configured to determine a value of a spectral index representing the biomass based on the one or more reflectance values. The processor is configured to calculate the biomass pigment surface area density based on one or more calibration values ​​determined for a reference coating and the spectral index SI. The processor is configured to calculate the biomass pigment surface area density by applying a reflectance compensation associated with the reflectance of the coating on the object.

[0050] Optionally, the system includes a digital camera. The digital camera can be used to acquire one or more images. The images acquired by the camera can be obtained via spectral bands. The camera can be, for example, a general-purpose digital camera, wherein one or more of the camera's red, green, and blue channels, and optionally an infrared channel, can be used as spectral bands. The digital camera may include spectral band filters, for example, having a bandwidth of 0-100 nm FWHM. The camera can be a multispectral camera. A multispectral camera can, for example, have a red spectral band from about 600 nm to about 700 nm. A multispectral camera can have an infrared spectral band from about 750 nm to about 950 nm or higher. A multispectral camera can, for example, have a blue spectral band from about 400 nm to about 500 nm. A multispectral camera can, for example, have a green spectral band from about 500 nm to about 600 nm. The camera can be a hyperspectral camera.

[0051] The digital camera can be held by a diver. Alternatively, the system includes a submarine, such as an unmanned submarine, for acquiring images of objects below the waterline. Alternatively or additionally, the system includes a hull crawler, an imaging unit that descends along the object, a (dry dock) photography chamber, etc.

[0052] According to another aspect of the invention, a computer program product is provided for estimating the biomass of a biofilm on a coating on an object permanently or intermittently immersed in an aquatic environment, comprising computer-implementable instructions that, when implemented by a programmable computer, cause the computer to:

[0053] - Obtain one or more digital images of contaminated portions of the coating on an object;

[0054] - Determine the corresponding reflectance value for the portion to be coated from each of one or more images; - Determine the value of the spectral index representing biomass based on one or more reflectance values;

[0055] - Calculate the biomass pigment surface area density based on the spectral index SI and one or more calibration values ​​determined for a reference coating; and

[0056] - The calculated biomass pigment surface area density is compensated for the reflectance of the coating on the object by applying a compensation associated with the difference in reflectance of the coating on the object relative to the reference coating.

[0057] Computer program products can be stored on non-volatile memory devices.

[0058] It should be understood that all the features and options mentioned in this method apply equally to system and computer program products, and vice versa. It will also be clear that any one or more of the aforementioned aspects, features, and options can be combined. Attached Figure Description

[0059] Embodiments of the present invention will now be described in detail with reference to the accompanying drawings, wherein:

[0060] Figure 1 An example of the system is shown;

[0061] Figure 2A and Figure 2B An example of the experimental setup is shown;

[0062] Figure 3 The reflectance spectrum of an exemplary coating is shown;

[0063] Figure 4 The reflectance spectrum is shown;

[0064] Figure 5A and 5B An example is shown where a definite value of the spectral index is used as a function to measure chlorophyll a density;

[0065] Figures 6A to 6F An example of measuring a spectrum is shown;

[0066] Figures 7A to 7F An example of a normalized spectrum is shown;

[0067] Figure 8 The NDVI estimation error is shown;

[0068] Figure 9A and 9B The estimates of the error factor and baseline error are shown;

[0069] Figure 10 The spectral bandwidth is shown;

[0070] Figure 11A and 11B The estimation error is shown; and

[0071] Figure 12A schematic representation of the method is shown. Detailed Implementation

[0072] Objects permanently or intermittently immersed in aquatic environments are susceptible to biofilm contamination. Such contaminating biofilms typically consist of a diverse community of microorganisms embedded in the extracellular matrix exudated from cells. The matrix provides attachment, adhesion, and protection. Examples of microorganisms typically responsible for biofilm matrix formation include bacteria, archaea, and microalgae such as cyanobacteria and protozoa. Because these types of microorganisms are key components, aquatic contaminating biofilms are interchangeably referred to as microfouling. However, contaminating biofilms may also include other transient elements incorporated into the matrix, such as debris or non-living sediments, embedded cells typically classified as planktonic organisms (e.g., phytoplankton), or small multicellular organisms such as fungi or small filamentous algae.

[0073] To reduce fouling, immersion surfaces (such as ship hulls, static oil rigs, and marine renewable equipment) can be coated with marine fouling control coatings that inhibit (but do not completely inhibit) biological growth.

[0074] Quantitative characterization of marine biofilms found on fouling control coatings provides insights into the coating's modes of action and efficacy. Microscopy, phospholipid analysis, microelectrodes, and optical coherence tomography are among the many techniques used to explore biofilm processes. A common thread across these methods is their reliance on the collection and / or analysis of small biofilm samples, which limits their practicality in capturing biofouling variations across large objects such as ships. Hyperspectral imaging has been found to be well-suited for mapping fouling by coloring organisms such as algae, microalgae, and / or bacteria.

[0075] Biofilms on contaminated objects (such as ships) are typically phototrophic. Diatoms are generally reported as the dominant microalgal taxa in contaminated biofilms. Other microalgal photobiobiota reported in contaminated biofilms include cyanobacteria and supplemental eukaryotes. The biomass of coloring bacteria and / or algae can be characterized by spectral reflectance data because algal taxa possess characteristic pigments (e.g., photosynthetic pigments and accessory pigments) that contribute to the characteristic reflectance spectral features. Chlorophyll a, also referred to herein as Chl a, is, for example, a pigment common to all photosynthetic groups, exhibiting strong absorption at approximately 670 nm, and this feature generally does not overlap with any other pigment peaks.

[0076] Spectral imaging captures quantitative spectral reflectance data and is therefore well-suited for examining large-scale spatial processes associated with algal biomass, such as those determined from the spectral properties of algae. Compared to sample-derived biofilm characterization methods, spectral imaging is easily scalable and thus has high potential for characterizing biofilm contamination processes that infiltrate objects. Images obtained through spectral imaging can include one or more pixels.

[0077] The advantage of this invention is that it is possible to map dirt over large areas to provide an overall picture of the structural dirt, and thus can help identify any problem areas that may require special attention, such as during cleaning. However, it can also be adapted for use over smaller areas, even individual pixels, allowing for faster access to information about specific areas of interest.

[0078] Hyperspectral imaging can quantify and map biomass using hyperspectral biomass indices calibrated based on algal pigment concentrations. Chlorophyll a is an established proxy for primary productivity and biomass, and is therefore suitable for hyperspectral calibration. Chlorophyll a can be indexed by comparing the depth of its absorption characteristic at approximately 670 nm with reflectance at reference wavelengths in the near-infrared or blue region, which are less affected by other pigments. Examples include the Normalized Differential Vegetation Index (NDVI) or the Benthic Microalgae Index (MPBI). Other indices for quantifying the biomass of Euglena (green algae) and diatoms (brown algae) have also been developed.

[0079] Biofilm reflectance spectra are modified from the spectrum of the underlying matrix. In marine transportation, colored marine fouling control coatings are fundamental; therefore, their spectral characteristics must be considered when performing spectroscopic measurements on biofilm fouling. For example, biocidal antifouling coatings are typically red or reddish-brown due to their high copper oxide content, coatings for floating marine renewable platforms require yellow for high visibility, and yacht antifouling coatings are formulated in iridescent colors. All of these colored coatings have non-flat reflectance spectra, thus methods can be used to ensure that the spectral reflectance of the background coating does not distort the experimental results.

[0080] Figure 12 A method 100 is shown for estimating the biomass of aquatic-derived biofilms on a coating of an object, for example, permanently or intermittently immersed in an aquatic environment. The method includes:

[0081] a) Obtain one or more digital images of the contaminated portion of the coating on the 102 object in the spectral band;

[0082] b) Determine the corresponding reflectance value for the portion to be coated from each of one or more images;

[0083] c) Determine the value of the spectral index representing biomass based on one or more reflectance values;

[0084] d) Calculate the pigment surface area density of 10⁸ biomass based on the spectral index SI and one or more calibration values; and

[0085] e) Compensate for the calculated biomass pigment surface area density by applying compensation associated with the reflectivity of the coating on the object.

[0086] The following section will describe examples of steps a) through e) in more detail.

[0087] Figure 1 An example of a system 10 is shown for estimating the biomass of aquatic-derived biofilm on a coating 12 on an object 14, for example, permanently or intermittently immersed in an aquatic environment. In this example, the object 14 is a vessel partially submerged below the waterline 16. In this example, system 10 includes a digital camera 18, such as a hyperspectral camera, arranged to obtain one or more digital images of the contaminated portion 20 of the coating on the object 14 within a spectral band. The spectral band may, for example, have a spectral width over a wide range of 0-100 nm FWHM (full width at half maximum). The digital camera may include one or more pixels. The digital camera may include a 1D sensor array, sometimes referred to as a line scan camera. The digital camera may include a 2D sensor array.

[0088] System 10 includes a processor 22 connected to or connectable to camera 18. The processor may be part of the underwater portion of the system. Alternatively, the processor may be part of the surface portion of the system. Processor 22 is configured to acquire one or more digital images from the camera of contaminated portions of the coating on the object. Processor 22 is configured to determine a corresponding reflectance value for the portion of the coating from each of the one or more images. Processor 22 is configured to determine a value representing a spectral index (SI) based on the one or more reflectance values. Processor 22 is configured to calculate the biomass pigment surface area density based on the spectral index SI and one or more calibration values ​​determined for a reference coating. Processor 22 is configured to compensate the calculated biomass pigment surface area density for the reflectance of the coating on the object by applying compensation associated with the difference in reflectance of the object's coating relative to the reference coating. This will be explained in more detail below.

[0089] The following experiments demonstrate that a good estimate of microfouling biomass can be achieved by compensating for the calculated biomass pigment surface area density of the coating on the object by applying a compensation associated with the reflectivity of the coating on the object.

[0090] Biomembrane experiment

[0091] The experiments were based on cultured microalgal and bacterial biofilms, closely resembling those observed in real-world marine microbial ship fouling. Monoculture biofilms of varying densities were prepared. Monoculture species were selected from algal communities reported in the literature as present in microfouling assemblages and / or observed in fouling biofilm samples. Mixed populations of microalgal and bacterial biofilms seeded from sampled marine fouling were also prepared and tested to provide insights into the impact of taxonomic heterogeneity.

[0092] The biofilm grows on glass, and the contaminated glass is positioned on each of the six coating colors for spectral imaging, allowing the same biofilm to be measured against different colored backgrounds.

[0093] Biofilm culture

[0094] Single cultures: Single cultures of two diatom species (Achnanthes sp., CCAP1095 / 1 and Diplophora sp., CCAP1001 / 3), green algae (Chlorella sp., CCAP211 / 53), and cyanobacteria (Arthrophora sp., CCAP1452 / 6) were obtained from the Scottish Association of Marine Science Culture Collection of Algae and Protozoa (SAMS CCAP). For each species, the original seed culture (20 mL) was scaled up to 3 x 125 mL cell suspensions at 18°C ​​with a 12-hour on / off light cycle (Sylvania T8 fluorescent white cold 840 + white warm 840 (Lorenz et al., 2005, Maintenance of Actively Metabolizing Microalgal Cultures. Algalculturing techniques, 145). Microalgae were cultured in nutrient-rich artificial marine microalgal growth media according to CCAP specifications: Guillard's f2+Si – a common marine microalgal growth medium with added silica to support diatom growth (Achnathes, Amphora), Guillard's f2 – without silica (Chlorella), (Guillard and Ryther, 1962, Studies of Marine Planktonic Diatoms: I. *Cyclocarya* nana Hustedt and Detonula). Confervacea (Cleve) Gran. Canadian Journal of Microbiology, 8, 229-239), or blue-green medium - for culturing cyanobacteria (Nodularia) (Stanier et al., 1971, Purification and characterization of single-celled cyanobacteria (Cyclocarya). Bacteriological Review, 35, 171).

[0095] Mixed population culture of *HP* (Hartlepool): The source population used for *HP* cultures was biofilm collected from non-toxic plates that were immersed at a depth of approximately 0.5 meters below the surface for 3–6 months at Hartlepool Marina on the northeast coast of England. Biofilm samples were mixed in Guillard's f2+Si medium to create a 500 mL cell suspension and filtered through a 125 μm filter to remove large particles.

[0096] Mixed-population culture of IP (International Paint): The source population used for IP culture was biofilm collected from a non-toxic plate that was immersed for one month at approximately 0.5 meters below the surface at Raffles Marina, Singapore. Biofilm samples were mixed in seawater to create a cell suspension and filtered through a 125 μm filter to remove large particles. 1 L of cell suspension culture was scaled up in International Paint’s 250 L circulating mixed-population autotrophic marine biofilm culture system at 25°5 with 12 hours of on / off light circulation (58 W Marine White, Photochemical Blue, Arcadia) (Longyear, 2014, Section 2 Mixed-population fermenters. Biofouling methods, 214). This system contained artificial seawater rich in Guillards f2+Si nutrient medium.

[0097] Biofilm formation

[0098] All biofilms grew on glass cover slips (circular, 19 mm diameter, Fisher Scientific). Due to their fragility, the cover slips were placed on glass microscope slides (25 x 75 mm, Fisher, 2-3 per slide) for support and secured with a microplate sealing film. The sealing film was pre-perforated with circular holes (9 mm in diameter) created by a hand punch, centered on the cover slip below. This configuration exposed only the central area of ​​each cover slip to contamination. When the masking film was removed after the growth period, the resulting biofilm had a well-defined perimeter and was surrounded by a contrast area of ​​clean glass.

[0099] For both monoculture and mixed culture of *Helicobacter pylori*, biofilms were grown in 4-well plates (Fisher Scientific) (2-3 plates per population). To cultivate a range of biofilm densities, the ratio of inoculum (mother culture) to nutrient medium added to each 10 mL well was increased sequentially, with the last well exclusively containing inoculum. The plates were sealed and the biofilm was cultured at 18°C ​​with a 12-hour on / off light cycle (Sylvania T8 fluorescent white cold 840 + white warm 840; (Lorenz et al., 2005, Maintenance of Actively Metabolizing Microalgal Cultures. Algal culturing techniques, 145). Every two weeks, 5 mL of nutrient medium per well was replaced with a serum pipette, taking care not to damage the biofilm, and the plates were resealed until biofilms of varying densities were visible across the masking membrane and the exposed glass. After the growth period (approximately 2 months), the mask was removed, and 8–12 slides with biofilms from each population were prepared for imaging.

[0100] For mixed-culture IP, a slide is placed horizontally in an illuminated, 8 cm deep channel integrated into the circulating culture system. Over several weeks, a spatially non-uniform biofilm occupies all surfaces of the channel, including the slide. The slide is removed from the channel, and the mask is removed as well. Cover slides contaminated with a range of biofilm densities and compositions are prepared for imaging.

[0101] Instrument Configuration

[0102] Biofilms were imaged using a hyperspectral line-scanning system (Resonon benchtop pika XC, 377-1029 nm, 3.3 nm band, 200 channels, 17 mm Schneider objectives, 30.8° field of view). Using this instrument, the sample was placed on an automated translation stage below the lens. The imager collected a single line of spectrum 1600 pixels wide and created an image by continuously capturing data at a user-defined frame rate as the sample stage moved across the field of view (FOV). The spatial width of each pixel was determined by the lens FOV and the distance from the lens to the imaging surface. The length of each pixel was determined by the speed of the linear translation stage and the frame rate. The imaging system was set to generate approximately square aspect ratio pixels of about 0.06 mm x 0.06 mm, which allowed imaging of the entire length of a 75 mm microscope slide within a single scan.

[0103] The translation stage is illuminated across the visible and near-infrared spectrum (ranging from approximately 380 nm to approximately 980 nm) by a four-beam assembly of a broad-spectrum quartz halogen lamp (Resonon). The instrument's exposure level is manually set using a lens iris to maximize the dynamic range of the image, allowing the reflectance spectrum measured from a white Teflon (Resonon) lamp to span across but not saturate the instrument's 14-bit sensitivity.

[0104] With the lens cap in place, the instrument's electrical noise spectrum was collected and interpreted using dark correction as specified in the manufacturer's instrument software (Spectranon). Variable illumination across the imaging field, attributable to lamp positioning, was measured by imaging a pure white Teflon panel and interpreted in the software using standard response correction as recommended by the manufacturer.

[0105] The spectral reflectance standard (50% flat spectral response from 360–1100 nm, NIST certified) is centrally included in the field of view of each experimental image to allow reflectance calibration during image processing.

[0106] Biomembrane imaging

[0107] Each biofilm is repeatedly imaged against a different colored background by gently repositioning a glass coverslip on a coated microscope slide.

[0108] Non-toxic background: Select dirt-releasing coating Intersleek TM 900 (International Paints Ltd.) offers grayscale (white, gray, black) and primary color (red, yellow, and blue) variants because they represent a range of brightness levels and spectral characteristics and are also non-toxic. Following the specified application protocol, the coating is applied by roller to a glass microscope slide, followed by an anti-corrosion primer and a dirt-relevant adhesive layer (Intershield). TM 300 and Intersleek TM 757). The slides were soaked in seawater at Hartlepool Marina for several months, then removed and thoroughly cleaned with a sponge.

[0109] Biofilm transfer and repeated imaging: A first color microscope slide is placed in an inverted well plate cover and filled with artificial seawater to a shallow depth (<2 mm). Three or four coverslips with biofilm are gently placed on each slide using forceps, taking care not to detach the biofilm. Images are collected when all the biofilm from each culture (8 to 12 biofilms per dish) is loaded into the dish. The next set of color slides is then placed adjacent to the first set, and the coverslips are gently slid from one color to another using forceps, after which the first color slide is removed. A second set of slides is centered on the dish, and a second image is collected, and so on. Approximately every two images, the water in the dish is replaced as it becomes warm under a halogen lamp. If any small fraction of the biomass is removed during transfer (although this is rarely observed), the order in which the colors are used is arbitrarily chosen to minimize the effect of any order. By using this strategy, spectra can be obtained from identical biofilms but with different colored backgrounds, thus limiting the background color to the only variable.

[0110] Biomembrane pigment analysis

[0111] Pigment analysis provides a traditional, quantitative, but destructive measure of microalgal biomass and generates a baseline dataset for post-processing spectral index calibration.

[0112] Pigment extraction: After biofilm imaging, the biofilm and coverslip were sandwiched in a glass fiber filter (0.45 μm), rapidly frozen, and stored at low temperature (LN2 gas phase) until extraction (8–10 months). Algal pigments were extracted by sonication (55W QSonica sonicator) for 1 minute in 1 mL of chilled methanol (MeOH, 100%, analytical grade) to rupture cell membranes and disrupt the biofilm matrix. Residual deposits were then removed by injection through an in-line 0.45 μm filter (Watman).

[0113] High Performance Liquid Chromatography (HPLC): HPLC analysis of biomembrane extracts followed a method outlined by Van Heukelem and Thomas (Van Heukelem and Thomas, 2001, Development of Computer-Aided High Performance Liquid Chromatography Applied to the Separation and Analysis of Phytoplankton Pigments. Journal of Chromatography A, 910, 31-49), adapted for use with an Agilent HP 1100 instrument (Agilent Technologies, CA) featuring a diode array detector (UV signal / wavelength = 450 nm / 4, wavelength range 350–750 nm) and a Zorbax Eclipse XDB-C8 (RP, 3.5 μm, 4.6 x 150 mm) column. On the day of analysis, prior to the run, up to 18 samples were first mixed with buffer (28 mM tetrabutylammonium acetate aqueous solution (TBAA), pH 6.5) at a 1:1 ratio, stored in amber bottles, queued in a refrigerated autosampler at 5°C, and run overnight. Mobile phases A (70:30 (v / v) methanol, 28 mM aqueous TBAA, pH 6.5) and B (methanol) were introduced at a flow rate of 1.1 ml / min according to the curve [0 min: 95% A, 5% B; 22 min: 5% A, 95% B; 31 min: 95% A, 5% B], with a column temperature of 60°C.

[0114] Determination of Pigment Area Density: The peak heights of 16 microalgal pigments, present in microalgal communities such as diatoms, green algae, cyanobacteria, dinoflagellates, and cryptobuds, were calibrated using HPLC instruments and methods, typically found in micropolluted boat and submerged plate samples. The pigments were chlorophyll a, b, and c2, polydinoflagellates, chlorophyll, amethyst, lutein, fucoxanthin, zeaxanthin, β-carotene, diadinoflagellates; xanthophyll; cyanobacterial lutein; divinylchlorophyll a, 19-hex-fucoxanthin; and neoxanthin; sourced from the commercial supplier DHI. Chromatograms of the experimental samples were analyzed, and peak identity was confirmed by UV absorption spectra of each peak associated with the calibrated pigments and by absolute and relative retention times. The concentrations of the sixteen calibrated pigments in each biofilm [ng / mL MeOH] were determined based on the chromatographic peak heights. The concentrations were converted to pigment surface area density [μg / cm³] by dividing each pigment concentration by the biofilm area. 2 The experimental images were scaled in ImageJ (NIH.1.51d) and then measured using the manual pixel selection tool.

[0115] Biomass estimation from hyperspectral indices

[0116] By developing calibration formulas, non-destructive hyperspectral measurements and destructive pigment analysis measurements are linked, converting spectral index values ​​into estimates of biomass.

[0117] Background coating reflectance spectrum

[0118] Figure 2A This illustrates how to measure the coating reflectance spectrum from a clean area of ​​the coverslip glass (solid line, 1); the biofilm spectrum from the region of interest (ROI), including all biomass (dashed line, 2); and paired local coating spectra from the ROI of the clean glass surrounding the biofilm (shaded area, 3). The ROI around the clean area of ​​the glass coverslip in the hyperspectral image is plotted using the manual selection tool of the imager's software package (Spectranon, version 2.68, Resonon, MT, USA) (see [link to image]). Figure 2A The average reflectance spectrum of the coating, Rcoating, was measured at point 1. The average reflectance spectrum of the 50% reflectance standard Rref was also measured from the hyperspectral image (see [reference]). Figure 2B (at 4 locations). By scaling the spectrum to a reflectance standard, the measured coating spectrum was converted from the digital luminance measurement [DN] to the percentage reflectance of incident light [0-100%], where

[0119] Rcoating[%]=Rcoating[DN]x0.5[%] / Rref[DN].

[0120] Figure 3The reflectance spectra of the white (W), gray (G), and black (Bk) coatings are shown to be relatively flat (64%, 24%, and 1% average reflectance, respectively), while the spectra of the primary color coatings yellow (Y), red (R), and blue (B1) exhibit color-specific reflectance and absorption characteristics.

[0121] In this experiment, a subset of gray (G) coatings from the hyperspectral images was selected as the reference coating for calibrating the hyperspectral index because the coating was found to have a flat spectral response and an overall reflectance that closely matches the previously published spectra of natural deposits. However, it will be understood that any other coating or substrate could be chosen as the reference.

[0122] Biomembrane reflectance spectrum

[0123] For each biofilm (n) of each culture (c), the average spectral Rbio of the ROI used to track the perimeter of the biofilm was measured. m(c,n) ( Figure 2A (arrow 2). The local gray coating Rgrey for each biofilm was also defined. m(c,n) ( Figure 2A Arrow 3) shows the area of ​​the clean glass cover glass surrounding the biofilm. The local coating spectrum was measured because the surface is not perfectly flat, leading to localized variations in illumination.

[0124] Biomass estimation using spectral indexing

[0125] Two hyperspectral indices extracted from microalgae remote sensing literature were evaluated for their suitability for estimating the biomass of biofilms in spectral variation experiments. The calibration of the pigment spatial density indices was ranked by their estimation errors.

[0126] Spectral indices: These indices combine wavelengths across the visible and near-infrared spectra. Table 1 below provides details on the hyperspectral indices of microalgal biomass.

[0127] In this paper, R673 is a defined reflectance value in the range around 673 nm, such as in the range of 605-740 nm (e.g., at 673 nm); R800 is a defined reflectance value in the range around 800 nm, such as in the range of 720-900 nm (e.g., at 800 nm); R673 clean It is the reflectance value of the coating on an object without biomass in the range of around 673 nm, such as in the range of 605-740 nm (such as at 673 nm).

[0128] Table 1

[0129]

[0130]

[0131] Exponential calibration of biomass: The power model fitted by nonlinear least squares estimation is defined as describing the spectral index Rbio calculated from the biofilm dataset. norm(g) The relationship between the spatial density of pigments in selected biomembranes, where

[0132] Index = a * pigment density b +c

[0133] Based on the fit, the calibration values ​​a, b, and c are determined.

[0134] The index values ​​of two indices, rNorm and NDVI, for all normalized biofilm spectra were calculated and compared with the biofilm density of the universal photosynthetic pigment chlorophyll a (a widely accepted biomass substitute).

[0135] Comparison of biomass estimation errors across exponential models: To quantitatively assess the effectiveness of exponential biomass estimation (measured by pigment surface area density), the power model was converted into a transformation formula, where...

[0136]

[0137] For each biomembrane, according to Rbio norm(g) The calculated index values ​​were converted into biomass estimates, and the sum of squared residuals and the mean estimation error (estimated - measured pigment density) were calculated for all indices.

[0138] Biomass estimation on any colored background

[0139] The dirt control coatings are colored in a range of standard and custom colors, each with potentially higher spectral reflectance and absorption characteristics than any dirt microalgae. The index calibration derived above was tested for its general applicability to micro-dirt across the six color coatings.

[0140] Color-specific coating normalized biofilm reflectance spectrum

[0141] The coating and biofilm spectra used for hyperspectral images of white, black, red, blue, and yellow coatings were processed according to the procedure outlined for the gray coating dataset: measuring biofilm and local coating spectra (Rbio m Rcoating m The data was converted to percentage reflectance, and all biofilm spectra were normalized to local coating measurements (Rbio). norm(w) Rbio norm(bk) Rbio norm(r) Rbio norm(bl) Rbionorm(y) The analysis also includes the original grey normalized data (Rbio). norm(g) ).

[0142] Biomass estimation on all coatings

[0143] The values ​​for the two spectral indices were calculated based on the coating-normalized biofilm reflectance spectra of all six coating colors and all biofilms. Biomass was then estimated from each index value for all biofilms on all coatings using a calibrated conversion formula.

[0144] The biomass estimation error (estimated minus measured pigment density) was calculated for each index and each color coating. To examine for any biomass-related estimation error patterns, the error for each coating color dataset was compared to the measured biomass, and a linear regression was fitted to determine the baseline error (intercept) and error factor (slope). The error factor and baseline estimation error for each coating color were compared to the minimum coating reflectance values ​​at the exponential wavelength. For NDVI, for example, the error factor and baseline error for each color coating were compared to the minimum coating reflectance at 673 and 800 nm. Logarithmic equations were calculated to describe the relationship between the baseline estimation error and the error factor and the coating reflectance at the exponential wavelength.

[0145] result

[0146] Biofilm formation

[0147] Overall, the experimental culture method successfully produced six groups of spatially confined biofilms with different densities.

[0148] Biomembrane pigments as determined by HPLC

[0149] Single culture: Across all species, chlorophyll a is the dominant pigment and is present at the highest density. For diatom species, fucoxanthin is also present at high levels, and significant minor pigments include chlorophyll c2, β-carotene, and, for *Diatomaceous*, diadinoxanthin. For *Arthrophyllus*, the minor pigment group is limited to β-carotene, but other additional phycobilin pigments that cannot be measured by HPLC may also be present and contribute to constructing biofilm reflectance spectra. *Chlorella* biofilms contain high levels of chlorophyll b, as well as detectable levels of minor pigments such as aurorasin, xanthophyll, zeaxanthin, lutein, and β-carotene.

[0150] Mixed culture: For HP biofilms, chl a is the main pigment, with a concentration of 1.6 μg / cm. 2The maximum density was given, providing a density range similar to most monocultures. Fucoxanthin was the next major pigment, confirming the observation of its abundance in these biofilms. Among the remaining group of pigments, chlorophyll b indicated the presence of green algae, while xanthin indicated the presence of cryptobud plants, and the high value of β-carotene was not easily attributed to any taxa.

[0151] Chlorophyll a density range for IP biofilms (maximum density 6 μg / cm³) 2 The diatoms were much larger than all other cultures, except for monocultures of the *Aspergillus* genus. Fucoxanthin was present in very small amounts, indicating that diatoms were not prevalent in these samples, consistent with microscopic examination. The xanthin indicating cryptobudding plants was highly concentrated in the redder biofilms, suggesting the identity of small, reddish-brown flagellated unicellular algae observed microscopically. Relatively high chlorophyll b levels indicated the presence of green algae, matching the bright limestone green visual appearance of some biofilms. The diverse additional pigments, including those that were unidentifiable, highlighted the diversity and heterogeneity of the IP biofilms.

[0152] Biomass estimation from hyperspectral indices

[0153] Microalgal biofilm reflectance spectrum

[0154] The measured reflectance spectra from six groups of laboratory biofilms exhibited different absorption characteristics, such as... Figure 4 As can be seen in the absorption spectrum shown, these characteristic absorption features are highly visible against the background flat spectral reflectance of the gray coating. Figure 4 The reflectance spectra of six laboratory biofilm groups are shown, with the y-axis indicating the observed reflectance range. Deeper spectral features indicate denser biomass. Panel (a) shows *Aspergillus* biofilm. (b) *Diatomum* biofilm. (c) HP mixed culture biofilm (d) *Arthrophyllus* biofilm. (e) *Chlorella* biofilm. Biofilm and (f) IP mixed culture biofilm. Significant noise exists in measurements below 400 nm, along with minimal illumination, and therefore wavelengths below 400 nm were not further considered in the experiments. In the infrared, the reflectance spectrum from 850 nm to 1100 nm shows almost no observable change except for increased noise, and therefore wavelengths above 850 nm were not further considered in the experiments.

[0155] The spectra of all six cultures were characterized by an absorption feature at 673 nm, attributable to chlorophyll a, which deepened with increasing biofilm density. Diatom biofilms showed an absorption feature at approximately 630 nm, a typical characteristic of brown algae absorption at the presence of chlorophyll c. Green algae biofilms were characterized by strong absorption from 400–500 nm, but almost no absorption near 500–600 nm. Cyanobacterial biofilms showed spectral absorption at approximately 625 nm, possibly attributable to phycocyanin. The spectra of the HP mixed population biofilm were very similar to those of the *Diatomaceous* genus, suggesting that diatoms are an important component of these biofilms. IP biofilms had inconsistent spectra that did not closely map to any single culture. The greener spectra of the biofilms were similar to those of the *Chlorella* genus, while the spectra of the reddish-brown biofilms were different.

[0156] Biomass estimation using spectral indexing

[0157] Figure 5A and Figure 5B The determined values ​​of the spectral indices rNorm and NDVI are shown, respectively, as the chlorophyll a density (in μg / cm³) measured from the normalized reflectance spectrum of the biofilm on the gray coating. 2 Functions in units of . Solid circles, squares, and triangles represent diatom samples (Diatoms, Aspergillus, Hartelp biofilm), while hollow circles, squares, and triangles represent green algae, cyanobacteria, and Singapore slime samples.

[0158] A biofilm cultured with the highest visual density of bright green IP (intracellular polymeric substances). Figure 5A , 5B The strong alignment of rNorm and NDVI index data for the highest density brown monoculture of *Cyclocarya* biofilm (hollow triangle) and solid square underscores that these two indices are indeed widely used tools for quantifying microalgal biomass.

[0159] Table 2 shows the following... Figure 5A and 5B The power model coefficients of the calibration curves shown (solid lines), calculated based on the normalized biofilm reflectance spectra of the gray coating, and the mean absolute error of these indices in biomass estimation compared with the extracted pigment density measurements, are shown.

[0160] Table 2

[0161]

[0162] Biomass estimation on any colored background

[0163] Biomembrane reflectance spectrum

[0164] Experimental measurement of spectra of biomembranes Rbio mIntensely shaped by the spectrum of the underlying coating ( Figures 6A-6F The relative reflectance of the biofilm and coating is more clearly shown in the normalized spectrum. Figures 7A-7F ), where each spectrum is normalized for the spectrum of the corresponding coating in the absence of a biofilm.

[0165] The spectrum of the background coating (see) Figure 3 The reflectance spectra measured from the experimental biofilms are clearly represented. Normalized spectra of biofilms on red, blue, and black coatings ( Figure 7B , 7D (and 7E) indicates that across a specific wavelength range of color, the biofilm is brighter than the coating (normalized reflectance > 1). Conversely, the biofilm is darker than the white, gray, and yellow coatings across the board (normalized reflectance < 1). Figure 7A , 7C And 7F), and the normalized spectra of the coatings on these colors of biofilm are very similar.

[0166] Although the intensity varies with the coating color, a strong chlorophyll a absorption feature near 673 nm can be clearly identified in the spectrum.

[0167] Clean (i.e., biofilm-free) gray, white, and black coatings exhibit very similar flat reflectance spectra, differing only slightly in brightness levels. Figure 3 The spectra of biofilms on the black coating show differences from those on gray or white coatings. For biofilms on the gray coating (this dataset is the basis for calibration in this experiment), the characteristics of the measured spectra closely match the characteristics of the normalized spectra; see [link to relevant documentation]. Figure 6C and 7C After normalization, all spectral values ​​were approximately equal to or less than 1, indicating that the biofilm was darker than the clean gray coating. The same general spectral characteristics were observed in the normalized spectra of biofilms on both the white and gray coatings. Before normalization ( Figure 6E (average reflectance 0-4%) and after normalization ( Figure 7E The spectral indication of the black coating (with a reflectance of 0.48–2.15) shows that the biofilm is more reflective than the black coating itself, except for those wavelengths where microalgae have strong absorption characteristics (e.g., about 673 nm).

[0168] The spectral reflectance characteristics of biofilms on red, blue, and yellow coatings are comparable to their white, gray, or black counterparts within wavelength ranges specific to each coating. The red dataset is divided into dark and bright spectral ranges, below and above approximately 550 nm, respectively. In the bluer wavelengths (<550 nm), the red coatings themselves reflect only weakly (see [link to relevant documentation]). Figure 3 In the spectral region of [the region], the biofilm is brighter than the coating, and its normalized spectral value is greater than 1.0 (maximum 1.98). Figure 7B The red coating is similar to the black dataset. In the red wavelength range (>550 nm), the brightest wavelengths are in the reddest region, with normalized spectral values ​​ranging from 0.18 to 1, and are similar to the white and gray datasets. For the blue coating dataset (…),… Figure 7D In the 400-500 nm wavelength range and the near-infrared (typical) range, spectral reflectance decreases with increasing biomass, with the blue coating being the brightest and similar to the gray and white datasets. The blue coating reflects very little light between 500 and 700 nm (approximately 5%, see [reference needed]). Figure 3 Within this wavelength range, the biofilm is generally brighter than the coating (normalized values ​​1–1.25), with the exception of a chlorophyll a absorption feature at 673 nm still present in the normalized spectrum, albeit fainter than in the white and gray datasets. The normalized spectrum of the yellow coating is similar in profile to that of the white and gray datasets and has similar spectral values ​​(0.15–1).

[0169] Biomass estimation error when applying gray calibration to other colors:

[0170] Density-specific goodness of fit: Figure 8 This shows the biomass estimation error (estimated value minus measured μg / cm³). 2 The color-specific estimation error (CVI) is due to the application of the NDVI calibration derived from the gray data to the other five colors, and is directional and specific to each coating color. The color-specific estimation error is directional and linearly correlated with biomass. For example, the error in the blue dataset (due to the biomass estimation error resulting from the application of the exponential calibration determined from the biofilm on the gray coating to the biofilm on the blue coating) consistently underestimates biomass, and the underestimation error increases linearly with increasing biomass (R0). 2 =0.84). For the blue-coated dataset, the baseline biomass estimation error and error factor (intercept and slope of the linear model, respectively) were significant (p<0.05, p<0.001). A similar pattern of biomass underestimation was also measured for the black-coated NDVI dataset (baseline error p<0.1, error factor p<0.001, R<0.84). 2 =0.853). Conversely, the NDVI values ​​in the white and yellow coated datasets led to an overestimation of the linear increase in biomass, although the slope of the linear model was significant only for the white dataset (P<0.05).

[0171] Spectral index biomass estimation error as a function of coating spectrum

[0172] The range of marine coating colors extends far beyond the six colors included in the experiment, and therefore, individual color patterns considering density-specific biomass estimation errors relative to the coating spectrum were used to infer a general relationship between coating color and hyperspectral index biomass estimation errors. The baseline estimation error BE and error factor EF for each color were compared to the minimum reflectance of the coating at wavelengths included in each index (e.g., for NDVI 673 or 800 nm). A non-linear relationship between coating spectral characteristics and estimation errors was evident.

[0173] Figure 9A and 9B Estimates of the error factor EF and baseline error BE for rNorm and NDVI are shown, respectively, and plotted for the minimum coating reflectivity at the wavelengths included in the exponent (673 nm for both).

[0174] Table 3 shows the models used to estimate the error factor EF and the baseline error BE, as shown from... Figure 9A and 9B It has been determined.

[0175] Table 3

[0176]

[0177] The baseline error and error factor for rNorm and NDVI biomass estimation across six coating datasets vary with the logarithm of the coating minimum reflectance (see [link to documentation]). Figure 9A , 9B The error factor relationship parameters are very similar for both exponents (for rNorm, NDVI are respectively: R...). 2 =0.99, 0.95, coefficients =0.26, 0.24). The rNorm baseline error increases logarithmically with the minimum reflectivity of the coating. Figure 9A R 2 =0.91, coefficient =0.086). NDVI error is not very significant on the logarithmic scale. Figure 9B R 2 =0.43, coefficient =0.025), and if desired, it can be described by a linear model.

[0178] Figure 10 Examples of possible different spectral ranges for digital images are shown, in which the reflectance spectrum relative to a Hartelp (HP) biofilm on a gray substrate is shown (reference). Figure 4 (Panel (e)). In this example, the first spectral range is selected around 673 nm, for example, centered at 673 nm. In this example, the second spectral range is selected around 800 nm, for example, centered at 800 nm. Figure 10In the diagram, w1a represents the first spectral range of 3.3 nm FWHM, and w2a represents the second spectral range of 3.3 nm FWHM. Figure 10 In the diagram, w1b represents the first spectral range of the 9.9 nm FWHM, and w2b represents the second spectral range of the 9.9 nm FWHM. Figure 10 In the diagram, w1c represents the first spectral range of 16.5 nm FWHM, and w2c represents the second spectral range of 16.5 nm FWHM. Figure 10 In the diagram, w1d represents the first spectral range of 29.7 nm FWHM, and w2d represents the second spectral range of 29.7 nm FWHM. Figure 10 In the diagram, w1e represents the first spectral range of 62.7 nm FWHM, and w2e represents the second spectral range of 62.7 nm FWHM. Figure 10 In this example, w1f represents the first spectral range at 100 nm FWHM, and w2f represents the second spectral range at 100 nm FWHM. In this example, the spectral band w1f is not centered at 673 nm. It will be understood that the spectral band w1f approximately corresponds to the red channel spectral R sensitivity of a commercially available digital camera. The spectral band w2f approximately corresponds to the infrared channel IR spectral sensitivity obtainable with a modified commercially available digital camera (e.g., after removing the IR cutoff filter).

[0179] Figure 11A and 11B The biomass estimation error (in μmaterial / cm³) is shown. 2 (in units), as a function of the bandwidth of the spectral band used. Figure 11A Related to NDVI. Figure 11B Related to rNorm. Each panel indicates the corresponding spectral bandwidth in nm. In each panel, the vertical axis represents the average estimation error (for each of the six colors) after applying compensation associated with the difference in reflectance of the object's coating relative to the reference coating. In each panel, the horizontal axis represents the average estimation error (for each of the six colors) without applying compensation associated with the difference in reflectance of the object's coating relative to the reference coating. Therefore, when the indicator point is located closer to the horizontal axis than it is located on the vertical axis, the compensation associated with the difference in reflectance of the object's coating relative to the reference coating reduces the biomass estimation error. As from Figure 11A It can be seen that, for NDVI, compensation associated with the difference in reflectivity of the object's coating relative to the reference coating reduces biomass estimation errors for all colors except black. (As shown from...) Figure 11BIt can be seen that rNorm, which is associated with the difference in reflectivity of the object's coating relative to the reference coating, reduces the biomass estimation error for all colors except black at larger bandwidth values.

[0180] While methods can be applied relative to any color of the object's current coating, they are sometimes less accurate when the coating is black. Therefore, when the object's current coating is not black, methods can be applied with improved accuracy, for example, when it has at least one wavelength in the 500 to 700 nm range, and preferably at least 5% reflectance for more than one wavelength, such as in one or more wavelengths or spectral bands used to determine the spectral index.

[0181] By increasing the error factor EF used to calculate rNorm and NDVI current and baseline error BE current The number of data points can improve the results for black coatings.

[0182] Based on the above, biomass should therefore be calibrated according to the reference spectral index. SI The calculated biomass estimate can be obtained through the error factor EF. current and baseline error BE current Adjustments are made for both, which are functions of the reflectance spectrum of the current coating where biomass exists. Then, the estimated biomass on the current coating is calculated. current This can be determined from the following equation:

[0183]

[0184] And use the error factor EF and baseline error BE, as shown in Table 3.

[0185] This can be combined with equations used to determine pigment surface area density based on spectral indices and calibration values ​​a, b, and c, for example, as found in Table 2.

[0186]

[0187] Two equations can be combined to form an equation.

[0188]

[0189] Therefore, it is possible to estimate the biofilm biomass originating from an aquatic environment, such as permanent or intermittent immersion in an aquatic environment, based on the spectral index SI and one or more calibration values, while compensating for the calculated biofilm pigment surface area density relative to the reflectance of the coating on the object by applying compensation associated with the reflectance of the coating on the object. Thus, biofilm biomass can be estimated on any colored object, such as a vessel, using the proposed equations and calibration values.

[0190] It will be understood that, in this paper, a gray coating with a reflectance of 24% is chosen as a reference point to determine calibration values ​​a, b, and c for estimating biomass pigment surface area density based on spectral indices. However, another reference coating, for example, with a different reflectance, could be chosen, which would change the values ​​of calibration values ​​a, b, and c. However, this does not affect the inventive concept behind this method.

[0191] It will also be understood that, in this paper, a gray coating with 24% reflectivity is chosen as a reference point to determine the error factor EF and baseline error BE, used to correct for different colored coatings beneath the micro-fouling layer. However, another reference coating could be chosen, for example, with a different reflectivity, which would change the values ​​of the error factor EF and baseline error BE. However, this does not affect the inventive concept behind this method.

[0192] Using this method, regardless of the color of the coating beneath the biofilm layer, a spectral index can be determined from one or more digital images. Based on this spectral index, the biomass pigment surface area density can be estimated as if the coating color corresponds to a reference coating color used to determine the calibration value. The estimated biomass pigment surface area density can thus compensate for the reflectivity of the object's current coating by applying a compensation associated with the reflectivity of the object's coating relative to the reference coating color as described above.

[0193] This method allows for the acquisition of one or more digital images of contaminated portions of an object's coating, such as underwater, using (unmanned) submarine vessels. The one or more images can include images at different spectral ranges, such as those with an FWHM of 100 nm or less, 50 nm or less, 20 nm or less, or 3 nm or less. One or more images can be acquired using a hyperspectral camera. Acquiring one or more images using a general-purpose digital camera is also possible.

[0194] Inspections by divers and ROV vessels can utilize commercial underwater cameras and lighting equipment. The cameras can be adapted with bandpass filters specifically tuned to the chlorophyll a red light absorption characteristics at 673 nm. If combined with suitable calibration and imaging protocols, incorporating rNorm biomass estimation into the underwater inspection is possible. NDVI imaging is performed at depths of approximately 2 m or more below the waterline. In embodiments, equipment such as IR lamp assemblies and IR-sensitive imaging systems can optionally be used to measure NIR reflectance from the surface.

[0195] In this document, the invention is described with reference to specific examples of embodiments thereof, which should not be construed as limiting the scope of the claimed invention. For the purpose of clarity and concise description, features are described herein as part of the same or separate embodiments; however, alternative embodiments having combinations of all or some of the features described in these separate embodiments are also contemplated.

[0196] In the example, a hyperspectral camera was used. It will be clear that it is also possible to combine a broadband digital camera with one or more bandpass filters. It is also possible to use a broadband digital camera in combined spectral band illumination. It is also possible to use a broadband digital camera, for example, using a red channel and optionally an infrared channel. The red channel can be from approximately 600 nm to 700 nm. The infrared channel can be from approximately 750 nm to 850 nm.

[0197] However, other modifications, variations, and substitutions are also possible. Therefore, the specifications, drawings, and examples should be considered illustrative rather than restrictive.

[0198] In the claims, any reference symbols placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of features or steps other than those listed in the claims. Furthermore, the words "a" and "an" should not be construed as limited to "only one," but rather are used to mean "at least one," and do not exclude a plurality.

Claims

1. A method for estimating a biofilm biomass originating from an aquatic environment on a coating of an object, comprising: a) obtaining one or more digital images of a contaminated portion of the coating on the object; b) determining, from each digital image of the one or more digital images, a respective reflectance value of the portion of the coating; c) determining, based on the respective reflectance values, a value of a spectral index representative of biomass; d) calculating a biomass pigment surface area density based on the spectral index SI and one or more calibration values determined for a reference coating, and e) applying a compensation to the calculated biomass pigment surface area density, the compensation being associated with a difference in reflectance of the coating of the object relative to the reference coating. The compensation is based on a comparative reflectance value representative of the coating underlying the biofilm biomass.

2. The method of claim 1, wherein, The comparative reflectance value is any one of:

3. The method of claim 2, wherein, - a reflectance value determined for a portion of the coating of the object from which all biomass has been removed; - a reflectance value determined from a reference object; or - a reflectance value stored in a database. Each digital image of the one or more digital images is obtained in a spectral band of 0 to 100 nm.

4. The method of claim 3, wherein, The comparative reflectance value is the minimum of the reflectance values in the one or more spectral bands for the coating underlying the biofilm biomass.

5. The method of claim 4, wherein, At least one of the spectral bands is selected to contain an absorption wavelength of chlorophyll and / or at least one of the spectral bands is selected to exclude an absorption wavelength of chlorophyll.

6. The method of any one of claims 1 to 5, wherein, In step e) the estimated biomass pigment surface area density biomass on the current coating is determined from the following equation: current biomass = (Spectral Index) * (Error Factor) * (Baseline Error) SI , the error factor EF current , and the baseline error BE current , 7. The method of any one of claims 4-5, wherein, At least one of the spectral bands is selected around 433, 460, 496, 555, 584, 601, 673 or 800 nm.

8. The method of any one of claims 4-5, wherein, The spectral index comprises a ratio of two reflectance values.

9. The method of any one of claims 1 to 5, wherein, The spectral index SI is one of:

10. The method of claim 6, wherein, - a coating normalized red reflectance value, defined as - a normalized difference vegetation index, defined as 12. The method of claim 11, wherein, wherein R673 is a reflectance value determined in the range of 605 to 740 nm; R800 is a reflectance value determined in the range of 720 to 900 nm; R673 clean is a reflectance value in the range of 605 to 740 nm at the coating of the subject without biomass.

11. The method of claim 10, wherein, R673 is the reflectance value determined at 673 nm, R800 is the reflectance value determined at 800 nm, and R673 clean is the reflectance value at 673 nm. - if the spectral index is the coating normalized red reflectance value, - if the spectral index is the normalized difference vegetation, Then EF current =0.26ln(R) current673 -1.00; BE current =0.0862ln(R) current673 +0.0164; and In step d), the pigment surface area density is determined according to the following equation, Then EF current =0.24ln(R) current673 -0.83; BE current =0.025ln(R) current673 +0.1025; wherein R current673 is the reflectance value of the current coating at 673 nm.

13. The method of claim 12, wherein, where a, b and c are calibration values.

14. The method of claim 13, wherein, - if the spectral index is the coating normalized red reflectance, then a = -1.921, b = 0.093, c = 2.477; and - if the spectral index is the normalized difference vegetation, then a = 0.618, b = 0.240, c = -0.

321. The one or more digital images are obtained using a submarine.

15. The method of any one of claims 1 to 5, wherein, The one or more digital images are obtained using an unmanned submarine.

16. The method of claim 15, wherein, 17. A system for estimating a biofilm biomass originating from an aquatic environment on a coating of an object, the system comprising a processor configured for: - obtaining one or more digital images of a contaminated portion of the coating on the object; ​ - determining, for each digital image of the one or more digital images, a respective reflectance value of the portion of the coating; - determining, based on the respective reflectance values, a value of a spectral index representative of biomass; - based on the spectral index SI and one or more calibration values determined for a reference coating, calculating a biomass pigment surface area density; and - applying a compensation to the calculated biomass pigment surface area density, the compensation being associated with a difference in reflectance of the coating of the object relative to the reference coating.

Images