Imaging method and device for measuring sunlight-induced chlorophyll fluorescence

Abstract

The invention relates to an imaging method for measuring sunlight-induced chlorophyll fluorescence (CF) of an object to be measured (30), in which the object to be measured is irradiated with sunlight (40), and light (41) reflected off the object to be measured, after being filtered to form filtered light (42) by means of an interference filter (15) that is tiltable vis-à-vis an optical axis, is recorded to form an image (20) by means of an image sensor (11) having several light-sensitive sensor elements (12) and is made available to a calculation and control device (16) by said image sensor, wherein an image series (21) of several recorded images of the object to be measured is created by means of the calculation and control device, said images being recorded at different tilt angles (δ, ε) of the interference filter in order to change, for a portion of the sensor elements within the image series and by means of the different tilt angles per image, an effective transmission wavelength (λ') of the interference filter at which said interference filter has maximum transmission, said effective transmission wavelength being related to the individual sensor elements, wherein the sunlight-induced chlorophyll fluorescence detected per sensor element is determined for this portion of the sensor elements by means of the calculation and control device on the basis of intensity values (IP, Px,y) of the filtered light (42) recorded by the respective sensor element per image of the image series; the invention furthermore relates to a device (10) for measuring sunlight-induced chlorophyll fluorescence of an object to be measured.

Description

[0001] Imaging method and device for measuring sunlight-induced chlorophyll fluorescence radiation

[0002] The invention relates to an imaging method for measuring sunlight-induced chlorophyll fluorescence radiation of a measurement object, for example, a plant, in which the measurement object is irradiated with sunlight and the light reflected from the measurement object is examined for the presence of sunlight-induced chlorophyll fluorescence radiation. The invention further relates to a device for measuring sunlight-induced chlorophyll fluorescence radiation of a measurement object, in particular a plant.

[0003] Photosynthesis is the biochemical process by which plants produce biomass (carbohydrates) from carbon dioxide and water. The process takes place in the leaf pigment chlorophyll, and the necessary energy comes from absorbed (sun)light. In addition to biomass, photosynthesis also produces oxygen, which is essential for most living organisms. Photosynthesis is therefore the most important metabolic process on Earth.

[0004] Absorbed sunlight that a plant cannot use for photosynthesis is re-emitted by chlorophyll, among other forms, as so-called fluorescence. This weak radiation in the red wavelength range (approximately 650 nm - 800 nm) is overwhelmed by the much stronger sunlight in daylight and is imperceptible to the eye. In principle, however, the light emitted by a plant consists of the sunlight reflected by the plant and the chlorophyll fluorescence. Chlorophyll fluorescence typically makes up between about 1% and about 5% of the light emitted by the plant at the corresponding wavelengths.

[0005] Chlorophyll fluorescence is a significant indicator of the efficiency of the photosynthesis process. The stronger the fluorescent light, the less efficient the photosynthesis, which in turn suggests problems within the plant. Therefore, chlorophyll fluorescence is an important measurement in plant physiology. Due to the low signal intensity of chlorophyll fluorescence against the strong background solar radiation, it can only be measured with sophisticated equipment to achieve a sufficiently good signal-to-noise ratio (SNR).

[0006] To facilitate the measurement of chlorophyll fluorescence, the effect is already being utilized that oxygen in the Earth's atmosphere strongly absorbs incoming sunlight in certain narrowband wavelength ranges, so that only a small fraction of the sunlight reaches the Earth's surface. One of these oxygen absorption lines lies at 760.7 nm (O₂-A) and has a width of just a few nanometers. Chlorophyll fluorescence is also emitted by the plant in this spectral range. Within the oxygen absorption line, the ratio of fluorescence light to sunlight is significantly more favorable, making the fluorescence light more easily detectable.

[0007] A method utilizing this effect to measure chlorophyll fluorescence is described, for example, in "A Snapshot Imaging System for the Measurement of Solar-Induced Chlorophyll Fluorescence — Addressing the Challenges of High-Performance Spectral Imaging," Kneer, C. et al., IEEE Sensors Journal, Vol. 23, No. 19, October 1, 2023. Of two highly sensitive, parallel cameras, one is equipped with a narrowband interference filter at 760.7 nm, measuring within the oxygen absorption line. The other camera measures with an interference filter at 757 nm outside the oxygen absorption line. The chlorophyll fluorescence in the recorded scene is to be determined from the two images. The described method requires two cameras and two interference filters. Furthermore, the cameras must have very high sensitivity (at least 16-bit grayscale resolution) in order to detect the weak chlorophyll fluorescence signal.

[0008] Alternatively, optical spectrometers can be used to determine chlorophyll fluorescence. Optical spectrometers measure at a single point and are not capable of achieving image-like spatial resolution. Only through a large number of sequential point measurements can a rough picture be built up, if at all. It is not possible to quickly image chlorophyll fluorescence in a scene using a spectrometer. Against this background, the invention aims to provide an imaging method and a device for measuring sunlight-induced chlorophyll fluorescence radiation with reduced technical complexity, low costs, and an improved signal-to-noise ratio (SNR).

[0009] This problem is solved by an imaging method having the features of claim 1 and by a device having the features of claim 12. Further particularly advantageous embodiments of the invention are disclosed in the respective dependent claims.

[0010] It should be noted that the features listed individually in the claims can be combined with one another in any technically meaningful way (even across category boundaries, for example between method and apparatus) and demonstrate further embodiments of the invention. The description further characterizes and specifies the invention, particularly in conjunction with the figures.

[0011] It should also be noted that the conjunction “and / or” used herein, which stands between two features and links them together, is always to be interpreted in such a way that in a first embodiment of the inventive object only the first feature may be present, in a second embodiment only the second feature may be present, and in a third embodiment both the first and the second feature may be present.

[0012] The terms "approximately" or "essentially" are intended to indicate a tolerance range that a person skilled in the art in this field would consider customary. In particular, the term "approximately" is to be understood as a tolerance range of the related quantity of a maximum of + / -20%, preferably a maximum of + / -10%.

[0013] The invention relates to an imaging method for measuring sunlight-induced chlorophyll fluorescence radiation of a measurement object, for example a plant, in which the measurement object is irradiated with sunlight and light reflected from the measurement object, after being filtered to filtered light by means of an interference filter that can be tilted relative to an optical axis, is captured by an image sensor having several light-sensitive sensor elements to form an image (hereinafter also referred to as a single image) and is provided by the image sensor to a computing and control unit, which may, for example, have a microprocessor, microcontroller, digital signal processor (DSP) or the like, as well as memory such as RAM, ROM, Flash, (E)EPROM and the like.

[0014] The processing and control unit creates an image series from several images of the object being measured, taken at different tilt angles of the interference filter. Tilting the interference filter to these different angles serves to change the effective transmission wavelength of the interference filter for at least some of the sensor elements within the image series. The transmission wavelength of the interference filter is the wavelength at which the filter exhibits maximum transmission.

[0015] By means of the computing and control unit, the sunlight-induced chlorophyll fluorescence radiation detected for each sensor element is determined based on tilt angle-dependent intensity values ​​of the reflected, filtered light recorded by the respective sensor element for each image of the image series, i.e., dependent on the effective transmission wavelength.

[0016] In other words, the sunlight reflected by the object being measured is filtered in such a way that, simply by tilting the interference filter, a spectral selection of the light falling on the respective sensor elements is achieved, depending on the filter's tilt angle. The individual images differ essentially only by the different tilt position of the interference filter. Only a single image sensor (i.e., one camera) and a single interference filter are required for chlorophyll fluorescence measurement. For the first time, it is thus possible to image and measure the chlorophyll fluorescence of objects, such as plants, using only one camera. By eliminating the need for a second camera and a second interference filter, both the costs and the technical complexity of a measurement system can be significantly reduced.At the same time, by creating the image series from different images with different tilt angles of the interference filter, several independent measurement results are obtained for each sensor element, which enable a significant improvement in the signal-to-noise ratio (SNR).

[0017] An imaging technique is a method that creates an image from measurements of a real object, whereby the measurement or information derived from it is spatially resolved and visualized using brightness values ​​or colors.

[0018] An image sensor is a device for capturing two-dimensional images of light using electrical or mechanical means. In most cases, semiconductor-based image sensors are used, capable of capturing light up to, for example, the mid-infrared range. The sensor elements of the image sensor can be arranged in a matrix. Each sensor element generates a pixel of the image. A pixel, also called a picture element, is a single intensity value in a digital raster graphic that corresponds to the image captured by the image sensor.

[0019] The optical axis is the axis of symmetry of a rotationally symmetric optical system. The symmetry of the surfaces is what matters, not their boundaries.

[0020] The image series (or image stack) comprises several individual images of the same scene captured by the image sensor. This means that the distance and orientation of the image sensor relative to the object being measured remain essentially unchanged during the creation of the image series.

[0021] When tilting the interference filter relative to the optical axis, the tilting or rotation axis is usually perpendicular to the optical axis.

[0022] According to an advantageous embodiment, the interference filter is used with a nominal transmission wavelength within an oxygen absorption range of sunlight, preferably within an oxygen absorption band at 687 nm (O2-B) and / or at 760 nm (O2-A), and with a sufficiently narrow bandwidth so that the intensity modulation described herein is achieved within the oxygen absorption range. Such an interference filter has a maximum bandwidth of approximately 12 nm, preferably a maximum of approximately 3 nm, and even more preferably a maximum of approximately 1 nm. With such a narrowband interference filter, inhomogeneous illumination of the image sensor can be achieved in the manner according to the invention, since even slightly inclined incident light rays relative to a surface normal of the interference filter are filtered by the interference filter with a transmission wavelength that is significantly shifted compared to the nominal transmission wavelength.

[0023] The inhomogeneity of the illumination of the image sensor can be further influenced by the opening or image angle of the optical system, which results from a focal length of an optics (e.g. lens, objective, mirror or similar) that projects the reflected light onto the image sensor and a size (e.g. diagonal) of the image sensor.

[0024] In further advantageous embodiments, a reference measurement is performed on a reference object that does not emit sunlight-induced chlorophyll fluorescence radiation. The reference object is irradiated with sunlight. Light reflected from the reference object, after being filtered by the tiltable interference filter, is captured by the image sensor to create a reference image and provided to the processing and control unit. The processing and control unit generates a reference image series from several reference images of the reference object, captured at different tilt angles of the interference filter (as during the initial image acquisition). Similar to the image series comprising the individual images of the measurement object in a measurement process, the reference image series comprises several individual reference images of the reference object.

[0025] The intensity values ​​of the reflected, filtered light of the object measured, recorded by each sensor element in the image series and dependent on the tilt angle or effective transmission wavelength, are assigned to corresponding reference intensity values ​​of the reflected, filtered light of the object, also dependent on the tilt angle or effective transmission wavelength, recorded by the respective sensor elements in each reference image of the reference image series. The sunlight-induced chlorophyll fluorescence radiation detected by each sensor element is then calculated as an intercept of a regression line determined by the processing and control unit between the tilt angle or effective transmission wavelength-dependent intensity values ​​of the respective sensor element in the image series and the tilt angle or effective transmission wavelength-dependent values ​​of the sensor element in the image series.The reference intensity values ​​of the relevant sensor element in the reference image series are determined based on the effective transmission wavelength. The intercept denotes the point where the regression line intersects one of the (coordinate) axes defined by the image series and the reference image series.

[0026] Preferably, the reference image series has the same number of reference images as the image series, so that a direct assignment of the reference intensity values ​​to the intensity values ​​of the image series is possible.

[0027] The use of linear regression leads to an improvement in the signal-to-noise ratio, so that a highly sensitive 16-bit camera can potentially be dispensed with, which also reduces the costs and technical effort for a measurement system.

[0028] In another advantageous embodiment, a reference measurement on a reference object that does not emit sunlight-induced chlorophyll fluorescence radiation is performed simultaneously with the measurement of sunlight-induced chlorophyll fluorescence radiation on the target object. Here, the reference object is positioned relative to the target object such that it is irradiated with sunlight. Light reflected from the reference object, after being filtered by the tiltable interference filter, is captured by the image sensor together with light reflected from the target object, also after being filtered by the tiltable interference filter. In other words, the light reflected from both the reference and the target object is filtered by the interference filter, and the reference object and the target object are imaged simultaneously on the image sensor.

[0029] In addition to the aforementioned set of sensor elements for which sunlight-induced chlorophyll fluorescence radiation is determined, another set of sensor elements is predetermined. These elements are used to record reference intensity values ​​of the filtered light from the reference object that depend on the effective transmission wavelength (i.e., tilt angle dependent). Furthermore, the effective transmission wavelength-dependent intensity values ​​of the filtered light from the measurement object, recorded by the respective sensor element of the first set for each image in the image series, are matched with corresponding reference intensity values ​​of the filtered light from the reference object, recorded by the sensor elements of the other predetermined set for each image in the image series.

[0030] Unlike the previously described reference measurement, where separate reference images are created for a separate reference image series and the assignment of the recorded reference intensity values ​​of the reference object to the recorded intensity values ​​of the measured object can be made via the position of the sensor elements on the image sensor, in the present embodiment, where the measured object and the reference object are recorded side by side in the same image, the assignment is made via the transmission wavelength effective for each sensor element during the creation of each image. That is, a recorded intensity value of the measured object with an effective transmission wavelength determined for the corresponding sensor element (of one part) is assigned a reference intensity value of the reference object of the sensor element (of the other part) for which the same transmission wavelength is effective during the creation of the image.Thus, each intensity value of the measured object in the same image is assigned a "correct" reference intensity value of the reference object, also contained in the same image. Finally, the sunlight-induced chlorophyll fluorescence radiation detected by each sensor element of one part is determined as an intercept of a regression line calculated by the processing and control unit between the intensity values ​​of the respective sensor element of one part of the image series (dependent on the effective transmission wavelength) and the reference intensity values ​​of the respective sensor elements of the other part of the image series (dependent on the corresponding effective transmission wavelength).

[0031] Even though, in the present embodiment, the reference intensity values ​​of the reference object are contained in the same image as the intensity values ​​of the measured object and are thus part of the image series of the measured object, the term "reference image" used herein can be understood analogously as that area of ​​the image in which the reference intensity values ​​of the reference object, recorded by the sensor elements of the other part of the image sensor, are depicted. Accordingly, a reference image series can likewise be understood as a series of images within this image area of ​​the image series.

[0032] According to a further advantageous embodiment, the reflectivity of the object being measured is determined as a slope of the regression line by means of the computing and control device.

[0033] In a particularly preferred embodiment, the reference intensity values ​​obtained from the reference measurement are used for several image series of the object being measured. That is, the reference measurement is performed independently of the measurement of the object being measured. For example, in some embodiments, the reference measurement can be performed only once to calibrate the image sensor, with all subsequent measurements referring to this reference measurement. Alternatively, the reference measurement can also be repeated at predetermined intervals, for example, to compensate for age-related changes in the properties (e.g., drift effects, defects, etc.) of the image sensor or its sensor elements. In yet other advantageous embodiments, an intensity of the sunlight not filtered by the interference filter, i.e., essentially the entire broadband incident sunlight, is recorded for each image in the image series.The solar spectrum is captured using a separate light sensor (e.g., an "incident light sensor"). Intensity fluctuations in sunlight between images in the image series are then compensated or corrected by the processing and control unit based on the captured intensity of the sunlight not filtered by the interference filter.

[0034] Further advantageous embodiments provide that for each reference image of the reference image series, the intensity of the sunlight not filtered by the interference filter is detected by means of the separate light sensor, wherein the intensity fluctuations of the sunlight between the reference images of the reference image series are eliminated by means of the computing and control device based on the detected intensity of the sunlight not filtered by the interference filter.

[0035] Particularly preferably, the intensity of the sunlight not filtered by the interference filter, recorded during the measurement of the object being measured, is normalized by means of the computing and control device to the intensity of the sunlight not filtered by the interference filter, recorded during the reference measurement of the reference object, in order to further improve the accuracy and robustness of the measurement of sunlight-induced chlorophyll fluorescence radiation.

[0036] The interference filter can be tilted about two tilting axes or about one tilting axis. Tilting the interference filter about only one tilting axis is particularly preferred, as this requires less technical effort. For example, the interference filter can be tilted relative to the optical axis such that a surface normal of the interference filter forms an angle of at most approximately ±45° with the optical axis, preferably at most approximately ±15°. All other tilting angles during the creation of the image and / or reference image series can be predefined between the specified maximum values. The tilting angles do not necessarily have to be equidistantly distributed within the specified angular range. The angular range can preferably be adapted to the aperture or image angle of the optical system used.

[0037] In further embodiments, the image series is created with 2 to 100 captured images, preferably with approximately 50 to 500 images. Between each image, the interference filter is tilted to a new angular position. The angular differences between the predetermined tilt angles can be constant or vary during the acquisition of the individual images in the image series. The latter is particularly advantageous when a spectral resolution is to be changed within a specific angular range, e.g., set higher or lower.

[0038] The invention further relates to a device for measuring sunlight-induced chlorophyll fluorescence radiation of a measurement object, in particular a plant, comprising an image sensor having several light-sensitive sensor elements, a lens for imaging light reflected from the measurement object onto the image sensor, an interference filter arranged between the measurement object and the image sensor and tiltable relative to an optical axis for filtering the reflected light to filtered light, and a computing and control unit for controlling the tilting of the interference filter and for receiving an image provided by the image sensor, wherein the computing and control unit is configured to perform an imaging method according to one of the embodiments disclosed herein.

[0039] It is understood that, with regard to device-related definitions as well as the effects and advantages of device-related features, full reference can be made to the disclosure of analogous definitions, effects, and advantages of the method according to the invention, and vice versa. A repetition of explanations of analogous features, their effects, and advantages can thus be omitted in favor of a more concise description, without such omissions being to be interpreted as a limitation of any of the disclosed subject matter of the invention. Further features and advantages of the invention will become apparent from the following description of exemplary embodiments of the invention, which are not to be understood as limiting and are explained in more detail below with reference to the drawing. In this drawing, the following are shown schematically:

[0040] Fig. 1 shows a side view of an exemplary measuring arrangement of an embodiment of a device according to the invention and a measuring object;

[0041] Fig. 2 shows a side view of a geometric relationship between light rays hitting the interference filter and the image sensor of the device from Fig. 1 in a first tilt angle position of the interference filter;

[0042] Fig. 3 shows in a side view the geometric relationship between the light rays hitting the interference filter from Fig. 2 and the image sensor of the device from Fig. 1 in a second tilt angle position of the interference filter;

[0043] Fig. 4 shows a perspective view of an interference filter that can be tilted about two tilting axes;

[0044] Fig. 5 is a diagram illustrating the effect of a shift in the transmission wavelength of the interference filter of the device from Fig. 1 as a result of tilting the filter;

[0045] Fig. 6 shows in an upper row contour plots of transmission wavelengths at three different tilt angles of the interference filter and in a lower row corresponding image simulations of the intensity of the light filtered by the interference filter in the respective tilt angle position;

[0046] Fig. 7 shows in an upper graph a brightness profile of an image element from images of an image series and in a lower graph a course of the transmission wavelength effective for this image element in the images of the image series;

[0047] Fig. 8 shows in an upper graph a brightness profile of an image element from reference images of a reference image series and in a lower graph a course of the transmission wavelength effective for this image element in the reference images of the reference image series;

[0048] Fig. 9 shows a comparison of the brightness profile of the image element of the image series from Fig. 7 with the brightness profile of the image element of the reference image series from Fig. 8, using five different images or reference images as examples;

[0049] Fig. 10 shows, in a left-hand representation, a brightness for the image element of the five images from Fig. 9 and, in a right-hand representation, a regression line created from pairs of measurements;

[0050] Fig. 11 shows a detailed representation of the relationship between a measurement on a measurement object and a measurement on a reference object and the resulting regression line;

[0051] Fig. 12 shows a strategy for determining chlorophyll fluorescence; and

[0052] Fig. 13 shows a brightness profile of an image element from reference images of a reference image series with a separately measured intensity of the incident sunlight and a brightness profile of an image element from images of an image series with a separately measured intensity of the incident sunlight.

[0053] In the different figures, functionally equivalent parts are always designated with the same reference symbols, so that they are generally only described once. The specific embodiments described below are purely illustrative in nature, serving to clarify the fundamental structures and processes.

[0054] Fig. 1 schematically shows a side view of an exemplary measuring arrangement of an embodiment of a device 10 for measuring sunlight-induced chlorophyll fluorescence radiation CF according to the invention and a measuring object 30, in this case, by way of example, a plant. The device 10 has an image sensor 11, which in turn has several light-sensitive, matrix-like arranged sensor elements 12, which can be seen by way of example in Fig. 4. The image sensor 11 of the device 10 is a component of a (digital) camera 13.

[0055] In the measuring arrangement shown in Fig. 1, the object being measured 30 is irradiated with sunlight 40.

[0056] As can be further seen in Fig. 1, the device 10 has a lens 14 for imaging light 41 reflected from the object 30 onto the image sensor 11, and also an interference filter 15 arranged between the object 30 and the image sensor 11, which can be tilted relative to an optical axis OA and is located in a beam path or beam cone 43 with focal point F of the lens 14 of the light 41 reflected from the object 30. The interference filter 15 filters the reflected light 41 into filtered light 42. A computing and control unit 16 of the device 10 serves to control the tilting of the interference filter 15 and to receive an image 20 provided by the image sensor 11 (see Fig. 12).

[0057] The interference filter 15 can be tilted either about one tilting axis 17, as shown in Fig. 1, or about two tilting axes 17 and 17' (see Fig. 4).

[0058] The interference filter 15 is preferably a narrowband filter with a nominal transmission wavelength within an oxygen absorption band at approximately 687 nm (O2-B) and / or at approximately 760 nm (O2-A), corresponding to a minimum of an oxygen absorption line of sunlight 40 (see Fig. 5). Narrowband is defined as a maximum full width at half maximum (FWHM) of the interference filter 15 of approximately 12 nm, preferably approximately 3 nm, and even more preferably approximately 1 nm. The filter 15 can be located in front of the lens 14, or between the lens 14 and the image sensor 11.

[0059] The computing and control unit 15 can, for example, comprise a microprocessor, microcontroller, digital signal processor (DSP), or similar device, as well as memory such as RAM, ROM, Flash, (E)EPROM, and the like (not shown individually). It is configured to perform an imaging method according to one of the embodiments disclosed herein, as explained in detail below.

[0060] Figure 1 also shows a reference object 35 with which a reference measurement can be carried out as described herein. The reference measurement can be performed as a separate measurement before or after one or more measurements on the measurement object 30.

[0061] The reference measurement can alternatively be performed simultaneously with the acquisition of the measurement object 30, in which case the reference object 35 is positioned during the acquisition of the images of the measurement object 30 such that it is imaged on the image sensor 11 simultaneously with the measurement object 30. Advantageously, the reference measurement can be performed with each acquisition of the measurement object 30, so that the illumination conditions, i.e., the incident broadband, unfiltered sunlight 40, are always identical for both measurements.

[0062] Fig. 2 shows in a side view a geometric relationship between light rays striking the interference filter 15 and the image sensor 11 of the device 10 from Fig. 1 in a first tilt angle position ö = 0° of the interference filter 15, and Fig. 3 shows the geometric relationship in an exemplary second tilt angle position ö = a / 2 of the interference filter 15, where a represents the maximum image angle.

[0063] The nominal transmission wavelength of the interference filter 15 applies only to perpendicular incidence of light. For non-perpendicular incidence of light, the transmission wavelength shifts towards shorter wavelengths X': with Xo the nominal transmission wavelength of the filter (here 760.7 nm), 9 the angle between light ray and surface normal nf (see Fig. 4) of the filter 15 and i the refractive index of the filter 15.

[0064] Even rays outside the optical axis OA strike the filter 15 at an oblique angle and thus have a shifted transmission wavelength X'. If the filter 15 is tilted further, the angle of incidence of the various light rays changes even more. Given the known focal length f of the lens and the size s of the image sensor 11, the effective transmission wavelength X' can be geometrically determined for each sensor element 12 (Fig. 4) of the image sensor 11. The maximum angle of view a is then given by:

[0065] Figures 2 and 3 show how the maximum field of view a of the camera 13 is obtained for a known focal length f of the lens 14 and size s of the image sensor 11. The angles of incidence 9 of the corresponding image rays for the individual sensor elements 12 (Fig. 4) of the image sensor 11 can be geometrically determined – even when the filter is tilted by the angle θ. In the drawings, the image sensor 11 is shown mirrored across the focal plane. This simplifies the geometric analysis. Figure 2 shows the interference filter 15 in a position not tilted relative to the optical axis OA, i.e., θ = 0°, while Figure 3 shows the filter 15 tilted by 6 = a / 2. At 5 = 0°, the marginal rays of the cone of rays 43 strike the filter 15 at an angle of incidence of 9 = ±a / 2. At θ = a / 2, the marginal rays of the cone of rays 43 strike the filter 15 at 9 = 0° and 9 = a.

[0066] In general, the tilting of the filter 15 can also be considered three-dimensionally about two tilting axes 17 and 17' and respective tilting angles θ and s. Then the angle of incidence 0 between image ray ps and filter 15 is obtained from the scalar product of image ray ps and surface normal nf of the filter 15, as shown in Fig. 4 by way of example for the interference filter 15 which can be tilted about two tilting axes 17 and 17'. The following applies:

[0067] It can be seen that the transmission wavelength changes continuously when the filter 15 is tilted. A transmission range 18 of the filter 15 shifts, for the filter 15 with the nominal transmission wavelength of 760.7 nm, across the left edge of the oxygen absorption line towards shorter wavelengths X, as shown in Fig. 5. Curve 44 represents the oxygen absorption in the depicted wavelength range from 700 nm to 800 nm as the relative transmission T. rAt 760.7 nm, significant oxygen absorption of sunlight occurs. Chlorophyll fluorescence (CF) also takes place in this region.

[0068] When the transmission area 18 of the filter 15 is located in the oxygen absorption line, little reflected light 41 reaches the image sensor 11. Conversely, when the sensor is located outside the oxygen absorption line, a large amount of reflected light 41 reaches the image sensor 11. The intensity of the reflected, filtered light 42 for each sensor element 12 thus changes depending on the tilt angle θ (and possibly s) of the filter 15.

[0069] Fig. 6 shows, in an upper row, contour plots of transmission wavelengths Ä' at three different tilt angles θ = -8°, θ = 0° and θ = +8° of the interference filter 15, and in a lower row, corresponding image simulations of the intensity of the light 42 filtered by the interference filter 15 at the respective tilt angle position. For the image simulation, a focal length of 35 mm for the lens 14 and a size of 1 inch for the image sensor 11 were assumed as examples, but these values ​​are not necessarily limited.

[0070] In the central representation (θ = 0°), filter 15 is positioned perpendicular to the optical axis OA. The upper contour plot shows that the transmission wavelength X' decreases from the center (760.7 nm) towards the edge. The image simulation below demonstrates that, due to oxygen absorption, little light is transmitted in the center, while the brightness increases towards the edge. This oxygen absorption creates a shadow, referred to here as the oxygen absorption shadow 45. When the filter is tilted (left θ = -8° and right θ = +8°), the transmission wavelengths X' shift upwards or downwards in the contour plots, respectively. The same applies to the oxygen absorption shadow 45 in the images below.

[0071] The measurement of chlorophyll fluorescence CF consists of stepwise changing the tilt angle θ (and optionally s) of the interference filter 15 and capturing an image 20 (see Fig. 12) for each angular position. During this process, the oxygen absorption shadow 45 moves across the image sensor 11, and the entire image area is scanned by tilting the filter 15 about one or both tilt axes 17 and 17', respectively. If the camera position and orientation remain constant and the scene is also static, the image content is the same in all images, while the spectral transmission of the individual sensor elements 12 changes. Several individual images 20 form an image series (see Fig. 12) across different tilt angles θ and s, in which spectral information is associated for each sensor element 12.

[0072] If a longer focal length lens 14 is selected, its aperture angle becomes smaller and the angle of incidence on the filter becomes shallower. The oxygen absorption shadow 45 becomes larger and can fill the entire image width of the image sensor 11. In this way, a scan around a single tilting axis 17 may be sufficient.

[0073] Fig. 7 shows, in an upper graph, a brightness profile 46 of an image element 22, whose maximum brightness H is normalized to 1, from images 20 of an image series 21 (see Fig. 12), and in a lower graph, a profile of the transmission wavelength X' effective for this image element 22 in the images 20 of the image series 21. For the image series 21, the brightness H (scale 0 to 1) is plotted against the image number n, which corresponds to a tilt angle θ, in the upper graph. In the lower graph, the geometrically calculated transmission wavelength X' is plotted against the image number n. The minimum of the brightness profile 46 can be identified as the minimum of the oxygen absorption line at 760.7 nm and used for calibration. Fig. 7 shows a typical profile.

[0074] If one considers the individual sensor elements 12 or the associated image elements 22, their brightness H (or intensity) changes depending on the tilt angle ö (possibly also s) of the filter 15, i.e. as the oxygen absorption shadow 45 moves across the sensor elements 12.

[0075] The brightness profiles exhibit a symmetrical dip. Along the respective flanks, the transmission range of filter 15 moves into or out of the oxygen absorption line (see Fig. 5). At the plateau of the dip, the transmission range of filter 15 lies within the oxygen absorption.

[0076] To determine the chlorophyll fluorescence CF, a reference object (not shown) of known reflectivity and without chlorophyll fluorescence emission is measured using device 10 (reference measurement) in addition to a measurement object (e.g., a plant). A measurement I is obtained from the reference object. Ref, which is proportional to the solar irradiance sin. For a sensor element 12 or image element 22 of the reference object, the brightness can be normalized to 1 if the reflectivity is known, i.e., sin = Δf ■ 1 / refl.

[0077] Fig. 8 shows, in an upper graph, a brightness profile 47 of a picture element 22, whose maximum brightness H is normalized to 1, from reference images 25 of a reference image series 26 (cf. Fig. 12), and in a lower graph, a course of the transmission wavelength X' effective for this picture element 22 in the reference images 25 of the reference image series 26. Between the total radiation P detected by the measurement object 30 ge The following relationship exists between s and the solar radiation Sin:

[0078] Here, Reflp is the reflectivity of the object being measured (e.g., plant) and CF is the chlorophyll fluorescence. All quantities are fundamentally wavelength-dependent, but in the narrow wavelength range around the oxygen absorption line, they can be considered constant, so a linear relationship can be assumed.

[0079] Fig. 9 shows a comparison of the brightness profile 46 of the image element 22 of the image series 21 from Fig. 7 with the brightness profile 47 of the image element 22 of the reference image series 46 from Fig. 8, using five different images 20 or reference images 25 as examples (markings). as shown in Fig. 9).

[0080] Substituting P from the above equation (2) ges When applied to sine, a straight line is obtained whose slope corresponds to the reflectivity Reft of the object being measured and whose y-intercept corresponds to the chlorophyll fluorescence CF in the wavelength range under consideration.

[0081] To determine the chlorophyll fluorescence CF, the intensity of the filtered light of the object 30 recorded by each sensor element 12 (or image element 22) is plotted against the intensity of the corresponding sensor element 12 (or image element 22) of a reference object, and a linear regression is performed. The more pairs of measurements are recorded at different intensities, the more accurate the result of the linear regression and thus the determination of the intercept (chlorophyll fluorescence CF) and the slope (reflectivity Refl of the object 30).

[0082] Fig. 10 illustrates this relationship. The left-hand diagram shows the brightness H (or intensity I) for image element 22 of the five images (labeled "l"-"5") from Fig. 9, and the right-hand diagram shows a regression line 27 generated from the pairs of measurements. The intensities of an image element 22 from a measurement on the object 30 are plotted against the intensities of an image element 22 from a reference measurement on the reference object. The more pairs of values ​​for different intensities are considered, the more accurate the linear regression is for determining the intercept corresponding to the chlorophyll fluorescence CF.

[0083] Fig. 11 shows a detailed representation to further clarify the relationship between a measurement 31 on the measurement object 30 and a measurement 32 on the reference object and the resulting regression line 27.

[0084] When measurements P and R on the measurement object 30 and reference object, respectively, plot the respective intensities I of sensor element 12 and image element 22 against the tilt angle cp (or image number n), a typical trough-shaped curve 46 and 47 (also referred to here as the trough curve) is obtained, which shows when the oxygen absorption shadow 45 has passed over the sensor element 22. In the measurement R of the reference object, an upper plateau 31 is found at high values ​​due to the high reflectivity, and a lower plateau 32 is found at low values ​​due to filter absorption. In the measurement P of the measurement object 30 (e.g., plant), the plateaus 31 and 32 are shifted relative to the reference object due to a different reflectivity and the additional chlorophyll fluorescence.

[0085] For evaluation, the measured values ​​of measurement P are plotted against the measured values ​​of the reference measurement R, as shown in the lower right of Fig. 11. This yields the regression line 27, whose slope corresponds to the reflectivity Refl of the measured object 30 and whose y-intercept corresponds to the chlorophyll fluorescence CF. It can be seen that the positions of the plateaus 31 and 32 essentially determine the course of the regression line 27. For the linear regression of the line's course, however, the measurement points at the transitions between the respective plateaus 31 and 32 ("flanks of the troughs") are relevant, as these distribute potential measurement errors across the entire line and thus reduce their influence. This leads to an improved signal-to-noise ratio (SNR) compared to a two-point measurement, in which measurements are taken only once inside and once outside the oxygen absorption line.

[0086] Fig. 12 illustrates a strategy for determining chlorophyll fluorescence CF in detail. For each tilt angle epi of the filter 15, an image 20 is taken, and a corresponding image series 21 is created for the entire measurement P on the test object 30. Similarly, for each tilt angle epi of the filter 15, a reference image 25 is taken, and a corresponding reference image series 26 is created for the entire reference measurement R on the reference object.

[0087] The trough curves 46, 47 are obtained when image series 21 and 26 are processed for a fixed sensor element 12 and a fixed image element coordinate (x,y), respectively. Rx,y( <pO für die Referenzmessung R und P x , y (cpi) for the measurement P on the object being measured as the intensity profile of the respective image element (x,y).

[0088] These intensity values ​​are inserted into the corresponding formula for linear regression to determine the slope and the y-intercept of the regression line 27, which correspond to the reflectivity Refl and the chlorophyll fluorescence CF, respectively.

[0089] The reference measurement does not necessarily have to be carried out in a separate additional measurement for every measurement on the object being measured.

[0090] It is conceivable that in measurement P on the object being measured, certain sensor elements or pixels are reserved for the reference measurement and that the reference measurement is carried out simultaneously with the measurement on the object being measured, i.e., with each capture of a single image, as has already been explained in detail in the general part of the description.

[0091] The image sensor can alternatively or additionally be pre-calibrated once with a reference measurement, and all further measurements, or a predetermined number of further measurements, on the object being measured can be referenced to this. Fig. 13 shows measured values ​​R of a pixel 22 from each of n (e.g., 70) reference images 25 of a reference image series 26, each with a separately measured intensity IS,R of the incident (direct) sunlight 40 during the measurement, i.e., the sunlight 40 not filtered by the interference filter 15, and measured values ​​P of a pixel 22 from images 20 of an image series 21, each with a separately measured intensity Is,p of the incident (direct) sunlight 40 during the measurement, i.e., the sunlight 40 not filtered by the interference filter 15. These intensities I s ,p, IS,R are detected by means of a separate light sensor 19 (see Fig. 1), which can be a so-called Incident Light Sensor.

[0092] Creating the image or reference image series 21, 26 can take several seconds in total, depending on the required exposure time of the camera. During this time, the solar radiation can change, e.g., due to clouds. Therefore, in further embodiments, the solar radiation during measurements P, R can be measured with the additional light sensor 19.

[0093] In the exemplary scenario shown in Fig. 13, it can be seen that the solar radiation I S R is relatively constant during the reference measurement R, i.e., during the creation of the reference image series 26, and only decreases slightly at the end. During the creation of the image series 21 on the measurement object 30, significantly stronger variations in solar irradiance I are observed. s ,p to identify factors that affect the measurement curve of the object being measured.

[0094] The influence of variations or fluctuations in solar radiation I S,R, IS,P during the creation of the reference image series or image series, the intensity values ​​recorded by the sensor elements 12 can be determined in the individual images of the respective series based on the recorded solar radiation I S ,R or Is,p should be corrected accordingly to prevent distortion of the measurement result due to fluctuations in the intensity of sunlight.

[0095] Furthermore, - additionally or alternatively - fluctuations in the intensity of sunlight between the times of creation of image series 21 and the reference image series 26 can be determined in a similar way using the recorded solar irradiance I. S ,R or I s ,p must be corrected to obtain comparable series.

[0096] The presented method makes it possible for the first time to measure the chlorophyll fluorescence of a subject, particularly plants, using only one camera (i.e., one image sensor) and one interference filter. Eliminating the need for a second camera and a second interference filter significantly reduces the cost of a measurement system.

[0097] By using linear regression for data analysis, the reflectivity of plants can be determined in addition to chlorophyll fluorescence. Furthermore, regression leads to an improved signal-to-noise ratio, eliminating the need for a highly sensitive 16-bit camera and thus further reducing the cost of a measurement system.

[0098] Reference symbol list

[0099] 10 Device

[0100] 11 Image sensor

[0101] 12 sensor elements

[0102] 13 Camera

[0103] 14 Lens

[0104] 15 interference filters

[0105] 16 Computing and control equipment

[0106] 17 First tilting axle

[0107] 17' Second tipping axle

[0108] 18 Transmission range

[0109] 19 Light sensor

[0110] 20 images

[0111] 21 image series

[0112] 22 Image element

[0113] 25 Reference image

[0114] 26 reference image series

[0115] 27 Regression line

[0116] 30. Object of measurement (e.g., plant)

[0117] 31 Upper Plateau

[0118] 32 Lower Plateau

[0119] 35 Reference object

[0120] 40 sunlight

[0121] 41 Reflected light

[0122] 42 Filtered reflected light

[0123] 43 Beam path / Beam cone

[0124] 44 Oxygen absorption

[0125] 45 Oxygen absorption shadows

[0126] 46 Brightness profile

[0127] 47 Brightness profile a Viewing angle ö, cp First tilt angle s Second tilt angle

[0128] X wavelength

[0129] 0 angle of incidence

[0130] CF Chlorophyll Fluorescence

[0131] F focus point f focal length

[0132] H brightness n number of images or reference images in an image series

[0133] H brightness

[0134] I Intensity

[0135] IP intensity measurement object

[0136] IR intensity reference object

[0137] Is,p Intensity of sunlight during measurement on the object being measured

[0138] IS,R Intensity of sunlight during measurement at the reference object nf Normal vector

[0139] O2-A oxygen absorption band at 760 nm

[0140] OA Optical Axis

[0141] P Measured values ​​of the object being measured (measurement) ps Image ray with respect to a sensor element

[0142] R Measured values ​​of the reference object (reference measurement)

[0143] Reflectivity s Size of the image sensor

[0144] T r Relative transmission x First spatial dimension y Second spatial dimension

Claims

- 27 - Patent claims 1. Imaging method for measuring sunlight-induced chlorophyll fluorescence radiation (CF) of a measurement object (30), in particular a plant, in which the measurement object (30) is irradiated with sunlight (40) and light (41) reflected from the measurement object (30), after being filtered to filtered light (42) by means of an interference filter (15) that can be tilted relative to an optical axis (OA), is recorded to form an image (20) by an image sensor (11) having several light-sensitive sensor elements (12) and is provided by the image sensor (11) to a computing and control unit (16), wherein by means of the computing and control unit (16) a series of images (21) is created from several of the recorded images (20) of the measurement object (30), which are recorded at different tilt angles (θ, s) of the interference filter (15) in order to be used to determine the different tilt angles (θ, s) of the measurement object (30).s) for each image (20) an effective transmission wavelength (X') of the interference filter (15) related to the individual sensor elements (12), at which it has maximum transmission, to change at least for a part of the sensor elements (12) within the image series (21), wherein by means of the computing and control device (16) for this part of the sensor elements (12) the sunlight-induced chlorophyll fluorescence radiation (CF) detected for each sensor element (12) is determined based on intensity values ​​(IP, P, ) recorded by the respective sensor element (12) for each image (20) of the image series (21) that depend on the effective transmission wavelength (X'). x , y ) of the filtered light (42) is determined.

2. Method according to claim 1, wherein the interference filter (15) is used with a nominal transmission wavelength within an oxygen absorption range of sunlight (40) and a half-width of at most about 12 nm, preferably at most about 3 nm and more preferably at most about 1 nm.

3. Method according to claim 1 or 2, wherein a reference measurement is taken on a reference object (35) that does not have sunlight-induced chlorophyll Fluorescence radiation (CF) is emitted, wherein the reference object (35) is irradiated with sunlight (40) and light (41) reflected from the reference object (35), after being filtered to filtered light (42) by means of the tiltable interference filter (15), is recorded by the image sensor (11) to form a reference image (25) and is provided by the image sensor (11) to the computing and control unit (16), wherein a reference image series (26) is created by means of the computing and control unit (16) from several reference images (25) of the reference object (35) which are recorded at the different tilt angles (8, E) of the interference filter (15), wherein the intensity values ​​(I) recorded by the respective sensor element (12) for each image (20) of the image series (21), which depend on the effective transmission wavelength (X'), are P , P x , y) of the filtered light (42) of the object being measured (30) each by the corresponding sensor elements (12) per reference image (25) of the reference image series (26) reference intensity values ​​(IR, R) dependent on the effective transmission wavelength (X') x , y ) of the filtered light (42) of the reference object (35) are assigned and the sunlight-induced chlorophyll fluorescence radiation (CF) detected per sensor element (12) is assigned as an intercept of a regression line (27) determined by means of the computing and control unit (15) between the intensity values ​​(IP, P) dependent on the effective transmission wavelength (X'). x , y ) of the relevant sensor element (12) of the image series (21) and the reference intensity values ​​(I , R) which depend on the effective transmission wavelength (X'). x , y ) of the relevant sensor element (12) of the reference image series (26) is determined.

4. A method according to claim 1 or 2, wherein a reference measurement on a reference object (35) that does not emit sunlight-induced chlorophyll fluorescence radiation (CF) is carried out simultaneously with the measurement of the sunlight-induced chlorophyll fluorescence radiation (CF) on the measurement object (30), wherein the reference object (35) is positioned relative to the measurement object (30) such that it is irradiated with sunlight (40) and light (41) reflected from the reference object (35), After being filtered to filtered light (42) by means of the tiltable interference filter (15), together with the light (41) reflected from the object (30), which is also filtered to filtered light (42) by means of the tiltable interference filter (15), the image (20) is captured by the image sensor (11), wherein, in addition to one part of the sensor elements (12) for which the sunlight-induced chlorophyll fluorescence radiation (CF) is determined, another part of the sensor elements (12) is predetermined, with which reference intensity values ​​(IR, R) dependent on the effective transmission wavelength (X') are determined. x , y ) of the filtered light (42) of the reference object (35) are recorded, with the respective sensor element (12) of one part being recorded for each image (20) of the image series (21) intensity values ​​(IP, P) dependent on the effective transmission wavelength (X') x , y) of the filtered light (42) of the object being measured (30) each corresponding to the effective transmission wavelength (X') recorded by the sensor elements (12) of the predetermined other part for each image (20) of the image series (21) x , y ) of the filtered light (42) of the reference object (35) are assigned and the sunlight-induced chlorophyll fluorescence radiation (CF) detected by each sensor element (12) of one part is assigned as an intercept of a regression line (27) determined by means of the computing and control unit (15) between the intensity values ​​(I) dependent on the effective transmission wavelength (X'). P , P x , y ) of the relevant sensor element (12) of one part of the image series (21) and the reference intensity values ​​(IR, R) dependent on the effective transmission wavelength (X'). x , y ) of the relevant sensor element (12) of the other part of the image series (21) is determined.

5. Method according to claim 3 or 4, wherein a reflectivity (Refl) of the object being measured (30) is determined by means of the computing and control device (16) as a slope of the regression line (27).

6. Method according to claim 3, wherein the reference intensity values ​​(I, R) obtained from the reference measurement x , y ) for several image series (21) of the measured object (30).

7. Method according to one of the preceding claims, wherein for each image (20) of the image series (21) an intensity (Is,p) of the sunlight (40) not filtered by means of the interference filter (15) is detected by means of a separate light sensor (19), wherein intensity fluctuations of the sunlight (40) between the images (20) of the image series (21) are compensated by means of the computing and control unit (16) based on the detected intensity (Is,p) of the sunlight (40) not filtered by means of the interference filter (15).

8. Method according to one of claims 3, 5 or 6 and 7, wherein for each reference image (25) of the reference image series (26) the intensity (IS, ) of the sunlight (40) not filtered by means of the interference filter (15) is detected by means of the separate light sensor (19), wherein the intensity fluctuations of the sunlight (40) between the reference images (25) of the reference image series (26) are eliminated by means of the computing and control unit (16) based on the detected intensity (IS, ) of the sunlight (40) not filtered by means of the interference filter (15).

9. Method according to claims 7 and 8, wherein the intensity (Is,p) of the sunlight (40) not filtered by the interference filter (15) detected during the measurement of the object (30) is normalized by means of the computing and control device (16) to the intensity (IS, ) of the sunlight (40) not filtered by the interference filter (15) detected during the reference measurement of the reference object (35).

10. Method according to one of the preceding claims, wherein the interference filter (15) is tilted about two tilting axes (17, 17') or about one tilting axis (17), preferably exclusively about one tilting axis (17). - 31 - 11. Method according to any of the preceding claims, wherein the image series (21) is created with 2 to 100 recorded images (20), preferably with 50 to 500 images (20).

12. Device (10) for measuring sunlight-induced chlorophyll Fluorescence radiation (CF) of a measurement object (30), in particular a plant, comprising an image sensor (11) having several light-sensitive sensor elements (12), a lens (14) for imaging light (41) reflected from the measurement object (30) onto the image sensor (11), an interference filter (15) arranged between the measurement object (30) and the image sensor (11) and tiltable relative to an optical axis (OA) for filtering the reflected light (41) to filtered light (42), and a computing and control unit (16) for controlling the tilting of the interference filter (15) and for receiving an image (20) provided by the image sensor (11), wherein the computing and control unit (16) is configured to perform an imaging method according to one of the preceding claims.

Images