Method for generating superimposed image and related image recording device

Abstract

The invention proposes a method and an associated image recording device (5) for generating a realistic superimposed image (1) which is true in terms of at least one false color FF (x, y) for visualizing a spectral image (3a, 3b), which method and image recording device enable the avoidance of an undesired color shift of such false color in the superimposed image (1), and despite the superimposition, the white light image (2) shown as the image background in the superimposed image (1) remains substantially high image brightness and the white light image (2) remains substantially the original image contrast. For this purpose, according to the invention, at least one color channel value of the white-light image (2), which color channel is complementary to the false color to be presented in each case locally, is reduced in each case locally in a targeted manner, and the intensity I (x, y) of the spectral image (3a) to be visualized also in the superimposed image is reduced as the intensity I (x, y) is higher at this point (x, y).

Description

Technical Field

[0001] The present invention relates to a method for generating a superimposed image, in which the superimposed image visualizes the image signal part of a white light image (WLI) and the image signal part of a spectral image together. Here, the spectral image can especially exist in the form of a fluorescence image, which thus visualizes fluorescence. In such a scenario, the local intensity value I(x, y) of the spectral image / fluorescence image is visualized in the superimposed image with at least one local false color FF(x, y).

[0002] Therefore, especially multiple different false colors can be used to visualize the spectral image in the superimposed image, and then the corresponding false colors allow inferring the underlying intensity of the original spectral image. And if only one single false color is used, for example, the saturation or brightness of the false color can allow inferring the intensity value I(x, y) of the spectral image.

[0003] The white light image (which should also be at least partially visualized in the superimposed image) can be visualized in the superimposed image, for example, by means of grayscale or by means of true color. It can also be stipulated that the local color channel values of the white light image (i.e., for example, the R / G / B values when using three color channels of red / green / blue) are mutually calculated with the relevant local signal values of the spectral image for generating the superimposed image. For example, the spectral image can exist here as a grayscale image, so that the signal value is the grayscale value (GW). Or, the spectral image already exists, for example, in a false color presentation, for example, when the spectral image exists as an RGB image in the false color presentation, the local signal value can be, for example, the R- / G- / B value.

[0004] The present invention also relates to an image recording device, which preferably can exist in the form of a medical visualization system / can exist as part of a medical visualization system. For example, the system can include an endoscope, an exoscope, or a microscope as the corresponding image recording device. In addition, the visualization system can also include other components, such as a camera control unit and a display device. The image recording device or the visualization system includes at least one image sensor, especially at least two image sensors, and the at least one image sensor / the at least two image sensors are set up for sensorially detecting the white light image and / or the spectral image (especially for sensorially detecting the fluorescence image). In addition, the device / system also includes an image signal processing unit, which is set up for calculating the superimposed image (especially as explained above) from the white light image and the spectral image (the images are respectively recorded by the at least one image sensor). The device / system can output the superimposed image here, for example, the superimposed image in the form of a digital superimposed image, and / or display it on the display device. Background Art

[0005] In the prior art, different types of fluorescence imaging have been previously known, in which a white light image and a fluorescence image are separately detected by sensing. In addition, there are also other imaging schemes, in which, in addition to the white light image, such spectral images are separately detected by sensing, and the spectral images only visualize a specific local part of the spectrum used for imaging. Therefore, in such a scheme, an image generation chain is constructed in terms of signal technology, which generates a white light image and separately generates a spectral image from the white light image. The white light image visualizes broadband wavelengths in the visible light wavelength range, while the spectral image visualizes a relatively narrow spectral range compared thereto, such as a spectral range in the near infrared (NIR). For example, if a fluorophore with an excitation wavelength under ultraviolet light is used, which emits a fluorescence wavelength under near infrared light, then, for example, a common color image sensor can be used and broadband illumination can be used to detect the white light image by sensing on the one hand, and in addition, an additional monochromatic sensor can be used to detect the near infrared fluorescence wavelength spatially separated therefrom as a spectral image / fluorescence image. In this case, the fluorescence image visualizes the intensity distribution I(x, y) of the fluorescence signal detected by the monochromatic sensor, and the fluorescence signal can, for example, be in a wavelength range with a bandwidth of less than 100 nm, and then the monochromatic fluorescence image is usually first recorded as a grayscale image.

[0006] In addition, it has been previously known to visualize such spectral images by means of false color rendering. For this purpose, a so-called mapping function is usually used, which assigns a specific false color to each intensity value I(x, y) of the spectral image or the resulting grayscale value GW, so that a colored false color spectral image can thus be obtained. It is even possible in this case to use only a single false color and then change the brightness and / or saturation of this false color in order to visualize different intensity levels in the spectral image.

[0007] In many applications, it is now advantageous to visualize the white light image and the spectral image for the user in the form of a superimposed image. One advantage of such a scheme lies in that the spatial impression of the scene recorded thereby can be visualized by means of the (ideally colored) white light image, and at the same time, for example, the fluorescence signal can become visible in the superimposed image by means of at least one false color of the spectral image.

[0008] There are also schemes that use a mapping function that maps the different intensity values visualized in the fluorescence image of the fluorescence signal to a predefined scale of different false colors. In this case, the fluorescence signal can be visualized by means of multiple false colors according to a predefined false color scale, for example, in the form of a so-called "heat map".

[0009] It is precisely in medical applications that the following problem often exists, namely, that individual image regions, especially those where the surface is wet and thus exhibits a high reflectivity, exceed the maximum presentable brightness range. Therefore, these image regions are visualized as white in the white light image, but the actual color information is basically lost because even white shows that all the used color channels of the used image sensor indicate the maximum intensity value, that is, a phenomenon is shown that is known by the concept of "clipping". However, this original clipping is not necessarily always disturbing, and it can even improve the spatial impression of the white light image. Therefore, in image signal processing, "clipping" is understood as exceeding the allowed and thus presentable signal range in the image, for example due to overshoot or overage, that is, the overflow of the corresponding signal. In the case of a white light image, this can be understood as a pixel or a larger image region in which all the used color channels exhibit signal full amplitude, that is, the corresponding possible maximum signal value.

[0010] If now the white light image and the fluorescence image are superimposed by simple signal addition, then what can occur is that individual image regions of the superimposed image now only exhibit additional clipping, in which no original clipping was previously observed. However, this additional clipping is very disadvantageous because the surgeon viewing the superimposed image can no longer draw reliable conclusions about the current intensity of the fluorescence signal in these image regions, and the intensity is precisely what should be visualized with false colors. Therefore, it is desirable, on the one hand, to maintain the original clipping as much as possible, especially in image regions where no obvious fluorescence signal is shown, but on the other hand, to avoid additional clipping in the superimposed image as much as possible, which only appears through the superimposition of the two images, so that the fluorescence signal can be visualized with the correct false colors in the superimposed image.

[0011] Another technical problem to be solved in the superimposed image is that precisely when the colored white light image is superimposed with the spectral image / fluorescence image presented in at least one false color, it can lead to an undesired color shift of the corresponding false color to be presented, especially in such color regions where, for example, the relevant fluorescence signal is detected by sensing, and the fluorescence signal should be highlighted in the superimposed image with false colors accordingly. This color shift is problematic in applications because the user should infer the intensity of the fluorescence signal detected by sensing based on the false colors. Therefore, the color shift can lead to a misinterpretation of the fluorescence signal intensity and should thus be at least limited or completely avoided. This is precisely necessary in medical applications because otherwise it can lead to incorrect diagnoses based on the incorrectly visualized false colors.

[0012] In the prior art, there have also been solutions for avoiding the "additional cropping" in the above-mentioned superimposed image and, if possible, the color shift occurring therein, namely the so-called "alpha blending": The solution of this method lies in globally (i.e., for the entire image) assigning corresponding weights to the respective pixel RGB values of the corresponding image data, i.e., for example, the white light image and the spectral image, and choosing the weights exactly such that the additional cropping in the resulting superimposed image is exactly avoided. Therefore, global transparency values are usually used in "alpha blending". Thus, even if there is no signal at all in the spectral image, the brightness of the WLI channel is reduced.

[0013] However, with this alpha blending, it is possible to ensure that in each color channel, the corresponding signal values of the superimposed image remain within the available / presentable value range. For example, if the entire white light image is weighted with a coefficient α1, the spectral image can be added after being weighted with a coefficient α2 = 1 - α1, so as to obtain the superimposed image therefrom. Although this solution avoids additional cropping, it has a decisive disadvantage, namely that the image brightness is also lost in the following image regions where this is not necessarily required. In other words, the superimposed image generated by this method often exhibits insufficient brightness of the white light background.

[0014] In addition, when generating the superimposed image, it is generally also desirable to avoid, as much as possible, contrast loss and also avoid loss of color saturation of the spectral image, i.e., in particular, loss of color saturation of the false color in which the spectral image is visualized in the superimposed image. Summary of the Invention

[0015] Starting from this background, the present invention is based on the following task: to propose an improved method that solves the above problems and can provide a true-color superimposed image with high color saturation and high contrast. In addition, the present invention should also provide a related image recording device with which such a high-quality superimposed image can be reliably generated, especially a high-quality superimposed image for medical applications.

[0016] According to the present invention, the features of method claim 1 are provided for solving the mentioned task. In particular, therefore, according to the present invention, it is proposed to locally and selectively reduce at least one complementary color channel value of the white light image relative to the remaining local color channel values of the white light image in order to solve the task in the method described at the beginning. This reduction does not occur globally (as in alpha blending), but directly or at least indirectly according to the relevant local intensity value I(x, y) of the spectral image. The complementary color channel value here is complementary to at least one local false color in which the local intensity value I(x, y) of the spectral image / fluorescent image (should) be visualized (as described at the beginning).

[0017] In other words, as described at the beginning, the spectral image is visualized in the superimposed image by means of at least one local false color, where local here means that the at least one false color can vary on the superimposed image according to the local intensity value I(x, y) of the original spectral image: for example, the color value (hue) of the false color can vary, or, for example, when only a single green false color is used, its saturation value or brightness value can vary. For a specific location (x, y) within the superimposed image and for the false color FF(x, y) to be visualized there, the complementary color complementary to this local false color can be determined at any time. Correspondingly, however, the complementary color and thus the complementary color channel values of the white light image can also vary on the superimposed image: if a blue false color is to be visualized at a first location, the red and green color channels of the white light image are complementary thereto at that location. And if a yellow false color is to be visualized at a second location, the blue channel of the white light image is complementary at that second location. Correspondingly, then, according to the method of the invention, the red and green color channel values (R values and G values) of the white light image are reduced at the first location, and the blue color channel value (B value) is reduced at the second location.

[0018] Thus, it is possible to determine separately (location-dependently) which of the color channels (usually three color channels R, G, B here) for presenting the white light image is complementary to the local false color. And subsequently the color channel value of this complementary color channel can be reduced accordingly locally, i.e., at the location of the local false color.

[0019] The technical effect achieved by the selective location-dependent reduction lies first in that the white light image is somewhat distorted in terms of its color fidelity. However, when the white light image is superimposed on the spectral image in the superimposed image, only those color portions of the white light image are emphasized thereby that correspond to the local false color to be presented locally according to the spectral image. And the complementary color of the white light image thereto, i.e., more precisely, the color channel values of the relevant at least one complementary color channel of the white light image are attenuated by the reduction, so that a significant color shift of the local false color to be presented in the superimposed image is exactly prevented during the superimposition. In other words, thus in this way and method it is ensured that the false color in the superimposed image is visualized in true color, i.e., without a significant color shift. Thereby, it becomes possible for the user to draw reliable conclusions about the underlying intensity of the original spectral image based on the visualized false color, which enables more accurate diagnosis especially in medical applications.

[0020] In the above method, it is preferred here that the remaining / other color channel values of the white light image (i.e., for example, R and G, if B is the complementary color channel value) are locally and selectively maintained or even increased directly or indirectly based on the local intensity value I(x, y) of the spectral image. By thus locally and purposefully increasing the other color channel values while decreasing the complementary color channel value, the image brightness can be locally at least approximately maintained respectively. Thereby, the contrast situation of the white light image can be better obtained, and thus a more realistic superimposed image can be obtained.

[0021] Furthermore, it is preferred in this method that the higher the corresponding local intensity value I(x, y) of the spectral image, the more strongly the at least one complementary color channel value of the white light image decreases position-dependently. Here, it also applies that this decrease does not necessarily have to occur directly based on the local intensity value I(x, y), but as will become even more apparent, for example, it can also occur based on false colors, which are themselves related to the local intensity value I(x, y) (the local intensity value is then indirectly related to I(x, y)). For example, the false color values presented by the false colors of the spectral image can also be used as input parameters to determine which color channel value of the white light image at a specific location (x, y) in the superimposed image is complementary to the local false color to be visualized at that location. It should be noted here that in image regions where the spectral image has no significant intensity, the risk of color deviation is smaller, so that only in those regions the complementary color channel values of the white light image need to be decreased very little or not at all. However, this is precisely advantageous for obtaining the color authenticity of these image regions of the white light image in the superimposed image. For this purpose, consider an image region in which the spectral image exhibits a zero intensity value: there, the white light image does not have to be adapted at all, that is, the complementary color channel values do not have to be decreased there either, so that the white light image can be visualized in these image regions in the superimposed image without any color change.

[0022] In summary, it can thus be said that, that is, according to the relevant local intensity value I(x, y) of the spectral image, the at least one complementary color channel value of the white light image can be decreased to different extents, especially not at all, at different positions (x, y) in the superimposed image.

[0023] It is also mentioned at this point that the method according to the invention can of course be used to generate a continuous video image data stream of a real-time superimposed image starting from a first real-time white light image stream and a second real-time spectral image stream. In this case, the intensity I(x, y) of the spectral image can vary over time and can also vary from one image to another depending on which color channel value of the white light image is reduced to what extent at which location. This is not comparable to alpha blending that varies over time, because alpha blending only globally changes the weights and does not change the weights in a position-resolved manner and at least indirectly based on the local I(x, y) of the spectral image.

[0024] It is also mentioned that, for example, the color channel values (R, G, B) of the false color representation of the spectral image can also be used as input parameters in the method to determine which color channel / color channel value of the white light image is complementary to the local false color at a specific location (x, y) of the white light image. In other words, the reduction of the complementary color channel value can be directly based on the color channel values of the false color spectral image to be visualized in the superimposed image. However, even in this case, there is at least an indirect correlation with the relevant local intensity I(x, y) of the spectral image, because precisely the false color of the false color spectral image is a function I(x, y) of the corresponding local intensity values of the original spectral image and is thus directly related to these intensity values.

[0025] Here, the term "local" can be understood in particular such that the local manipulation of the corresponding color channel value of the white light image is carried out at a specific location (x, y) within the superimposed image, more precisely based on the corresponding intensity value I(x, y) at the same location in the spectral image / fluorescence image. It goes without saying that for this purpose, the white light image and the spectral image should have the same size in the desired superimposition. If this is not initially the case, for example due to different position resolutions of the images, the same image size can be achieved by appropriately scaling at least one of the two images.

[0026] Using the method according to the present invention, it is possible to achieve, for example, that at those positions (x, y) in the white light image (2) where high intensity values I(x, y) are sensed and detected in the corresponding spectral image (3a), the local color channel values (R / G / B) of the white light image (2) belonging to these positions are adapted such that the following color channel values (R / G / B) of the white light image (2) are reduced, and these color channel values do not correspond to the relevant local signal values (R / G / B) of the false color rendering FF(x, y) of the spectral image at these positions (x, y), that is, they are complementary to the corresponding false colors. Thereby, it is possible to effectively prevent a relevant color shift (color drift) of the false color to be presented in the resulting superimposed image when the white light image and the false color rendering of the spectral image are superimposed.

[0027] Therefore, the main technical advantage of the method according to the present invention is that the desired false color rendering of the spectral image is retained, and the desired false color rendering can be associated with, for example, an intensity false color scale. It should be noted here that the superposition / overlay of the same or similar color parts does not result in a relevant "color drift", but only results in the superposition of the false color with its corresponding complementary color. However, since the method suppresses these complementary colors locally (by targeted position-resolved / local reduction of the complementary color channel values of the white light image according to the corresponding intensity I(x, y) of the spectral image, for example, the "B value" in the yellow false color) in the relevant image area (i.e., where the relevant intensity I(x,y) is observed in the spectral image), the "color drift" of the corresponding false color to be presented in the superimposed image can be very effectively avoided. If only a single false color (such as false colors with different saturations and / or brightnesses) is used to visualize the intensity of the spectral image (such as from light green to dark green), or if multiple false colors (such as a continuous false color scale from black through dark blue, cyan, green to yellow) are used and thereby - especially only - the intensity I(x, y) of the spectral image is visualized by changing the color values (hues) of the false colors.

[0028] For this purpose, the HSV color space (H = hue = color value; S = saturation = color saturation; V = value = brightness) is observed exemplarily: for example, the false color rendering can only change the brightness V for specific H values; or the brightness and saturation remain constant, and the hue value changes according to the intensity I(x, y) (= moving along an arc path in the HSV color space).

[0029] In both cases, a very realistic and bright superimposed image can be obtained from the result. The white light part in the low-intensity I(x, y) region of the spectral image in this superimposed image can actually be reproduced almost error-free (because there is no significant adaptation of the image signal part of the white light image there), while the false color rendering of the intensity distribution of the original spectral image in the superimposed image is true color in the sense that the user can still reliably and accurately infer the original intensity I(x, y) in the spectral image based on the visualized false color and / or its saturation and / or its brightness. Exactly for the users who use this method in medical imaging, the image perception is thus greatly improved, and a more accurate interpretation of the superimposed image can be achieved. For example, this can lead to improved diagnosis.

[0030] Viewed from a slightly different perspective, the solution according to the invention can also be described as follows:

[0031] To solve this task, it can be stipulated that - in particular supplementary to the method features explained above and / or supplementary to the method according to claim 1 - the superimposed image visualizes the image signal part of the white light image and the image signal part of the spectral image, in particular the fluorescence image, together. To solve this task, it can also be stipulated that, in order to generate the superimposed image, the corresponding local color channel values (e.g., R / G / B) of the white light image and the relevant corresponding local signal values (e.g., R / G / B or GB) of the spectral image are mutually calculated using a corresponding ratio selected position-dependently and directly or indirectly according to the corresponding original local intensity value I(x, y) of the corresponding pixel (x, y) of the spectral image. This ratio (which can be understood as a type of local weight between the white light image and the spectral image) can therefore preferably be selected / changed pixel by pixel, more precisely especially as described above, directly or at least indirectly according to the local intensity value of the spectral image.

[0032] With this solution, the corresponding weight between the local color channel values of the white light image and the local signal values of the spectral image can thus vary position-dependently in the superimposed image, i.e., according to the local original intensity value I(x,y) of the spectral image. This variation can here be selected pixel by pixel or at least individually for specific image regions / can be varied pixel by pixel or region by region.

[0033] According to the invention, therefore (unlike in global alpha blending), appropriate weights between the image signal portion of the white light image and the image signal portion of the spectral image / fluorescent image can act in the superimposed image depending on the position, respectively, at a specific point in time. In the solution according to the invention, the local weights can be changed here according to the currently detected intensity distribution I(x, y) of the spectral image, so that the weight distribution on the superimposed image can also change from image recording time to image recording time. Thereby, the method according to the invention allows, for example when generating a video image data stream (composed of successively recomputed superimposed images based on a corresponding first white light image video stream and a second spectral image video stream), a dynamic adaptation of the superimposition process in terms of time and space. Here, the dynamics are generated by temporal and / or spatial changes in the spectral image intensity I(x, y), since the temporal and / or spatial changes in the spectral image intensity affect the corresponding local weights locally applied during the superimposition of the two images.

[0034] The original spectral image can here have corresponding signal values / pixel values that correspond to the correspondingly sensed intensity I(x, y) of the optical signal of the spectrum, in particular of the fluorescent signal. However, as an intermediate step, a false color rendering (false color image = FCI) of the spectral image can also be generated in a manner known per se. If a "heat map" is used for this purpose as the mapping function (which thus maps the corresponding intensity I(x, y) into the false color space (this false color subspace can only cover a single color value or, for example, cover a continuous range of color values / hue values, as explained for the HSV color space before)), then the false color image (FCI) can also have different false color signal values, i.e., in particular different false color / color values / hue values. Thereby, the signal values of the spectral image can basically exist as intensity values I(x,y), or can already exist as false color signal values (such as R / G / B values) after appropriate image processing.

[0035] For example, the color channel values of the white light image (WLI) can be: R / G / B values, i.e., the signal values of the corresponding color channels (= color signal values).

[0036] Viewed from a slightly different perspective, the solution according to the invention can also be described as follows:

[0037] In a method for generating a superimposed image, which method can be designed in particular as described above and / or according to the features of claim 1 and / or according to the features of claim 2, it is provided that the superimposed image jointly visualizes image signal portions of a white light image and image signal portions of a spectral image, in particular a fluorescence image. In addition, in order to solve this problem, it can be provided that, for generating the superimposed image, the corresponding ratios by which the local color channel values (R / G / B) of the white light image and the associated local signal values (R / G / B; GW) of the spectral image are calculated are selected in a position-dependent manner, preferably pixel by pixel, to generate the superimposed image. In order to solve this problem, it can be provided that the corresponding ratios are calculated using a superimposition function OF(I)=f(I(x, y)) (superimposition function), which directly or indirectly specifies, as a function of the corresponding local intensity values I(x, y) of the original spectral image, by which the corresponding local color channel values of the white light image are increased / enhanced / emphasized or reduced / weakened / reduced in the calculation for generating the superimposed image. The superposition function can therefore be understood as a “superposition function” which specifies with which respective local weightings the two images are to be superimposed on one another in different image regions of the superposition image.

[0038] Thus, the overlay function OF(I) can in particular determine which color channel values (e.g. R / G / B) of the white light image should be attenuated as complementary color channel values (B) relative to a specific local false color FF(x, y) to be presented in the overlay image, and / or which of these color channel values of the white light image should be enhanced or at least not attenuated as non-complementary color channel values. With this approach, the local false color FF(x, y) to be presented can also be presented in the overlay image in true color, i.e. without a noticeable color shift. In any case, a color shift is classified as noticeable if it leads to a noticeable error in the conclusion about the intensity I(x, y) underlying the false color for the spectral image.

[0039] For example, if the superposition function OF(I) is to produce an RGB representation of the superposition image, it can be specified as a vector OF(I)=(ΔR(I), ΔG(I), ΔB(I)) consisting of three corresponding correction functions. Here, for example, ΔG(I) is a correction function which specifies by which value the color value G of the green color channel of the superposition image is to be increased or decreased depending on the intensity I(x, y) of the spectral image. Technically equivalent to this, the superposition function OF(I) can also be obtained with the aid of corresponding correction functions (for each color channel, i.e., for example, for red: ΔR=f(R FF ,G FF ,B FF)) to form, and this correction function does not consider the intensity I(x, y) of the spectral image, but rather the R of the false color rendering FF(x, y) of the spectral image FF - / G FF - / B FF values, where these false colors FF(x, y)=f(I(x, y)) are related to the intensity I(x, y) of the spectral image, for example

[0040] ΔR=f(R FF (I(x, y)),G FF (I(x, y)),B FF (I(x, y)))

[0041] where, for example, G FF (I(x, y)) is the false color value of the green color channel of the false color spectral image (FCI), which is calculated position-dependently from the local intensity I(x, y) of the spectral image respectively.

[0042] The above three methods (the above three methods can also be used in combination to solve the above task) can also be extended as follows:

[0043] For example, it can be stipulated that when calculating the superimposed image, the false color rendering FF(x, y) of the spectral image is generated in the form of a false color spectral image from the original spectral image by means of a mapping function MF(I). This step can in particular be an intermediate step when calculating the superimposed image. In addition, the false color rendering or the false color spectral image can also be only partially visualized in the superimposed image. This is because, for example, in image regions where the original spectral image does not exhibit a significant intensity, the false color rendering can, if possible, completely disappear, so that in these image regions of the superimposed image, only the original white light image - without any changes if possible - is visualized. When using such a mapping function MF(I), it is preferred that: the mapping function assigns a specific local false color FF(x, y) to the corresponding local intensity value I(x, y) of the original spectral image. Thus, a mathematical mapping between the range of possible intensity values detected at each position in the spectral image (for example, in the case of an 8-bit resolution with a numerical range of 0...255) and the false color space (in which the corresponding false colors are encoded), such as the HSV color space, is described.

[0044] The mapping function MF(I) can here be mapped into a false-color subspace (e.g., within the mentioned HSV color space). This false-color subspace (which is a subset of the available color space) can in particular include only a single unique false-color value H, but can for example include different color saturation values S and / or different color brightness values H. For example, if the intensity of a spectral image is to be presented with false-color green, it is sufficient to change the color saturation value S of this false color in the false-color spectral image in order to thereby visualize the intensity distribution in the original spectral image.

[0045] However, the false-color subspace can also include a plurality of different false-color values H. For example, false colors from dark blue via light blue, cyan, green to yellow can be used in order to visualize the intensity of the original spectral image in a "heat map". In this case, it is preferred that these different false-color values form a continuous path within the false-color subspace. This is because in this case the continuous intensity distribution of the original spectral image can be visualized by the continuous color change within the false-color subspace in the superimposed image. This is advantageous for being able to infer the original intensity I(x,y) as accurately as possible from the superimposed image.

[0046] Accordingly, a preferred design provides that a visualization system, for example a visualization system in the form of an endoscope or a microscope (which can in particular be designed as described here), is designed such that the visualization system is set up for the sensor-based detection of a white-light image (WLI) and a spectral image (i.e., in particular a fluorescence image) and for visualizing a superimposed image. Here, the superimposed image visualizes the image signal portion of the white-light image and the image signal portion of the spectral image together. The visualization system is also set up for visualizing the superimposed image using an overall color space, where the white-light image is visualized using the colors of the VIS color space as a subset of the overall color space, while the spectral image is visualized using the false colors of the remaining residual color space of the overall color space (i.e., the VIS color space and the residual color space complement each other to form the overall color space used). In addition, the false colors of the residual color space are selected disjointly from the colors of the VIS color space; that is, the false colors are precisely not part of the VIS color space, but are only located in the residual color space. The size of the VIS color space can here also be changed according to the current VIS image and the tissue visualized in the VIS image (e.g., if initially adipose tissue is not visible and then yellow adipose tissue becomes visible, which should be visualized in the VIS image). It is thus characterized in that the false-color scale (or the color bandwidth of the false colors used) never penetrates into the VIS color space (i.e., is designed disjointly from this VIS color space).

[0047] This solution can be understood as follows: the VIS image is reproduced by the visualization system in a color rendering that occupies a specific part of the available color space, i.e., for example, the color spectrum from green to red. Accordingly, there is a residual color space, i.e., the remaining remainder of the entire available color space. The present invention recognizes here that it is advantageous for visualization that the false colors used for visualizing spectral images / fluorescent images do not intersect with the part of the color space used for presenting the VIS image. This is because, for example, the location where fluorescence appears in the scene can be accurately identified thereby.

[0048] Preferably, the false color scale can even be designed complementarily to the VIS color space, i.e., occupy the entire remaining residual color space. This is advantageous because the false color scale can then have the maximum length / color bandwidth, which in turn is beneficial for being able to resolve as many intensity levels of the fluorescent signal as finely as possible with different colors. For example, if only the color space from cyan to yellow is used, perhaps three colors, i.e., yellow, orange, and green, can be recognized by the user. However, in the solution proposed here, the false color space can extend from dark blue via cyan, green, orange to yellow, so that five different false colors and thus five intensity levels can be distinguished by the surgeon on the screen.

[0049] In particular, it can be provided that the spectral image is visualized by means of a false color scale (color map) that extends continuously within the residual color space. Preferably, the false color scale extends over at least 80% of the remaining residual color space here. This is because the intensity differences of the fluorescent signal can be visualized by means of a wide palette of false colors from the residual color space thereby. Here, ideally, the false color scale is displayed to the user together with the VIS image and / or the superimposed image, whereby the user can easily infer the intensity of the visualized fluorescent signal according to the scale.

[0050] The visualization system can also be set up to continuously determine the current VIS color space (and thus also automatically determine the remaining residual color space) based on the image information of the VIS image detected by sensing, and can be set up to determine the false color scale according to the determined current VIS color space. For example, if the color bandwidth in the VIS color space detected currently by sensing (using an image sensor) increases, for example, because previously invisible yellow adipose tissue now appears in the field of view, the visualization system can accordingly make the color bandwidth of the residual space smaller, i.e., transfer, for example, the yellow that is a false color from the residual color space to the VIS color space. Such an approach can always achieve an optimized false color rendering with the largest possible color bandwidth. According to this solution, the bandwidth of the residual color space for false color rendering is dynamically adapted according to which color bandwidth is exactly needed for presenting the VIS signal in the VIS color space.

[0051] For example, if yellow adipose tissue is visualized in the VIS image, it is advantageous if the false color scale no longer extends into the yellow area. It is completely particularly advantageous if the visualization system itself continuously / durably determines the occupied VIS color space from the current VIS image (which is detected by the image sensor of the system in a sensing manner), and thereby determines the residual color space that is always available at present. Then, the visualization system can automatically adapt the false color scale such that the false color scale always lies within the currently determined residual color space. Thereby, different intensity levels and / or different local signal values of the spectral image can be visualized to the surgeon with as fine a color gradation as possible using the false color scale.

[0052] The above-mentioned mapping function assigns a specific false color to the intensity as explained. The mapping function can be designed such that this assignment / mapping is carried out for all image points with a false color representation FF(x, y), where the false color representation is calculated from the intensity distribution I(x, y) of the spectral image by means of the mapping function: MF(x, y)=f(I(x, y)).

[0053] For example, if the mapping function is supposed to produce an RGB false color representation, the mapping function can be described as a vector consisting of three individual mapping functions for the respective color channel values (R / G / B), i.e., for example as MF(I)=(R(I),G(I),B(I)), where R(I) is the mapping function for the red color channel, which assigns a specific color value R(I) of the red color channel to each intensity value I(x, y).

[0054] The first alternative with only a single unique false color value mentioned above can be realized, for example, by means of a mapping function based on a brightness scale and / or a saturation scale of a specific false color (such as green). The second alternative (i.e., in the case of using different false color values) can be realized, for example, by means of a mapping function based on a preferably continuous false color scale. Thus, the false color scale can include a continuous spectrum of different color values. In the second alternative, the false color spectral image can thereby exist, for example, in the form of a color "heat map"; in the first alternative, the false color spectral image can exist as a brightness image in a single green false color.

[0055] The false-color subspace (residual color space) can be selected in a specific manner and method, in particular based on the color subspace (VIS color space) occupied by the white-light image (WLI). Generally, that is, the false-color subspace only covers a part of the color space that is generally available for presenting the superimposed image; the remaining part of the color space can be occupied by the WLI image without color overlap, which is advantageous for the reliable color interpretation of the superimposed image. This will be explained in more detail below.

[0056] At least "partial" visualization here can be understood as that only a specific local part of the false-color spectral image can be visualized in the superimposed image, or for example, a specific region of the false-color spectral image can be hidden in the superimposed image.

[0057] The false-color spectral image or false-color presentation FF(x, y) preferably can visualize false-color blue, cyan, green, and yellow, preferably in this order. Thus, this can be particularly advantageous in medical applications because it enables the visualization of different intensity levels of the original spectral image in a precisely distinguishable color.

[0058] Therefore, the mapping function MF(x, y) can in particular describe the continuous false-color change in the false-color subspace, preferably from blue to cyan to green to yellow. Here, it is particularly preferred that a high intensity value of the spectral image is assigned to the yellow false color, and a low intensity value of the spectral image is assigned to the blue false color. This is because this conforms to the color intensity inference that is intuitively correct for humans. The false-color spectral image can also include black as a false color here, that is, especially in the image regions where the intensity value of the spectral image is lower than the intensity threshold.

[0059] The false-color spectral image or false-color presentation FF(x, y) can also, for example, only visualize the false-color green and, if possible, also visualize the false-color black. That is, thus, the complex intensity distribution of the original spectral image can also be visualized in the superimposed image. Therefore, the mapping function MF(x, y) can describe the continuous brightness change and / or saturation change in the false-color subspace for at least one false color, especially for the false-color change from black to green. Here, it is preferred that a high intensity value of the spectral image is assigned to the green false color, and a low intensity value of the spectral image is assigned to the black false color.

[0060] In the method according to the invention, it can be provided that the local color channel values of the white-light image and the corresponding false-color signal values of the false-color spectral image are mutually settled with corresponding different position weights / local weights, that is, mutually settled according to the local original intensity values of the spectral image. This scheme can be particularly applied to the position-related settlement mentioned above.

[0061] As already mentioned, the corresponding local false colors of the false-color spectral image (i.e., the false colors that should be locally used to visualize the intensity of the original spectral image in the superimposed image) can define the corresponding local complementary color KF(x, y). In this case, it can be stipulated that the mentioned at least one color channel value of the white-light image (which corresponds to the complementary color KF(x, y)) is reduced position-dependently - in particular pixel by pixel. With this scheme, at least the color shift of the corresponding local false color in the resulting superimposed image can be restricted or completely prevented, which color shift can occur due to the settlement of the white-light image and the spectral image.

[0062] In all the above-mentioned schemes, which are technically equivalent according to the invention, the described reduction of the complementary color channel values and / or the corresponding settlement and / or the described ratio are not carried out directly according to the intensity value I(x, y), but only according to the local false color FF(x, y) to be presented. Thus, the false color FF(x, y) to be presented - more precisely its color value H and / or its saturation S and / or its brightness V - is a function of the original local intensity I(x, y) of the spectral image (since the false color should visualize the intensity distribution of the possibly monochromatic original spectral image / fluorescence image). In this scheme, the superimposition type is also carried out position-resolvedly at least indirectly according to the corresponding intensity value I(x, y) of the spectral image. This also applies if the false-color presentation FF(x, y) is a non-linear function of I(x, y), because even then it is still always related to I(x, y).

[0063] Of course, the method according to the invention can also be applied only to a plurality of individual image regions. That is to say, it is not necessarily always necessary to take into account the entire part of the spectrally imaged image that has been detected in a sensing manner, but for example the superimposition can be restricted to such an image region in which a distinct and thus to-be-presented intensity has been completely detected in a sensing manner in the original spectral image.

[0064] As already explained, the superimposed image can be calculated based on a superimposition function OF(x, y), which can be understood as a "superimposition function". In particular, this superimposition function can be predefined here and / or can be adapted by the user (e.g., via the operating interface of the image recording device used). The superimposition function now directly or at least indirectly prescribes the amount by which the corresponding local color channel values of the white light image should be increased / enhanced / emphasized or decreased / attenuated / reduced according to the corresponding local intensity values I(x, y) of the original spectral image. Here, it is preferred that when the local color channel value of the white light image corresponds to the relevant local false color (which should be implemented as the image signal part of the spectral image at the location (x, y) in the superimposed image), the local color channel value of the white light image is increased / enhanced / emphasized. Additionally, it can be advantageous that when the local color channel value of the white light image is complementary to the relevant local false color (which should be implemented as the image signal part of the spectral image at the location (x, y) in the superimposed image), the local color channel value of the white light image is decreased / attenuated / reduced.

[0065] As already mentioned at the beginning, both the white light image and the spectral image / fluorescence image can be sensed and detected using a single image recording device, where this device can in particular be designed in the form of an endoscope, an exoscope, or a microscope, and will be explained in more detail later. However, here the device can include multiple image sensors.

[0066] In the case of the sensed and separate detection of these two images, various scenarios can be envisaged: For example, the white light image and the spectral image / fluorescence image can

[0067] i) be sensed and detected separately from each other spatially, in particular by means of at least two separate image sensors and / or by means of different color filters at the pixel level of an image sensor (in particular an RGBX sensor or a hyperspectral sensor); or

[0068] ii) the sensed and separate detection can occur separately from each other in time, in particular by means of time-varying illumination. This time-separated detection can also be carried out based on the time-alternating detection of the two images using one (possibly the only) image sensor.

[0069] iii) The separate sensed detection of the two images can also be achieved only by intelligent signal processing, in particular, here the two images can be sensed and detected simultaneously and / or by only one single image sensor.

[0070] All these scenarios for separating the spectral image / fluorescence image from the white light image are basically already known from the prior art; however, they can be advantageously used in the solution according to the present invention for generating high-quality superimposed images.

[0071] As mentioned, only a part of the spectral image, i.e., a specific local image part of the spectral image, can also be visually presented in false colors in the superimposed image.

[0072] In contrast, the corresponding part of the white light image can be visually presented as a colored white light background image in the superimposed image, preferably in true color or as a monochromatic gray-scale background image. Even in the gray-scale image, the advantages of the present invention are retained, i.e., no additional cropping occurs. Even the "cropped" VIS / WLI image part is still colored.

[0073] If the white light image is visually presented in color in the superimposed image, the white light image can accordingly include color information of at least two different color channels. However, the white light image can also be visually presented in the form of a gray-scale image.

[0074] In contrast, the original spectral image can visually present the intensity values of the spectrally detected light signals, especially fluorescence signals, sensed (using an image sensor). If the present invention is used in the minimum solution, the white light image does not necessarily have to be visually presented in color in the superimposed image, and furthermore, the spectral image can be presented, for example, in the visualization image with only a single false color.

[0075] A particularly preferred design of the method provides that, when calculating the superimposed image, it is checked position-resolvedly - i.e., especially pixel by pixel - whether, by superimposing the image signal part of the white light image with the image signal part of the spectral image, the value range available for the superimposed image and / or the maximum presentable value in the superimposed image is exceeded. If this should be the case, the image signal part of the white light image, especially the at least one complementary color channel value mentioned, can be reduced to such an extent that the value range is locally maintained. With this scheme, it is precisely possible to avoid: additional cropping in the superimposed image, as described at the beginning and often observed in the prior art.

[0076] It is mentioned at this point that the original cropped-image region of the white light image - which has already shown to exceed the presentable value range (especially in the form of white image spots) - is still fully presentable in the superimposed image. This especially applies to image regions in which the original spectral image has negligible intensity and is accordingly less visible until not visible at all. These original cropped-image regions can possibly be visually presented in the correct false colors in the superimposed image, as long as the spectral image has a significant intensity corresponding to the respective original cropped-image region, so that the false colors are accordingly to be visually presented in this image region.

[0077] In short, this solution is used to avoid the appearance of additional "white spots" in the superimposed image in image regions such as the following, where the relevant intensity I(x, y) of the spectral image should be visualized using false colors. However, false colors can only be correctly visualized if not all (locally) color channels used for visualizing the superimposed image have the maximum presentable signal value (which corresponds to the color white in the superimposed image). Thus, with this solution, the original / current clipping that already exists in the recorded white-light image can be maintained in the superimposed image because individual pixels of the image sensor used exhibit signal full amplitude. However, the method according to the invention ensures that no obvious "color drift" of the image signal part of the spectral image visualized using false colors occurs in the superimposed image. At the same time, additional clipping can also be avoided, which can only occur through the superposition of two images and thus cause the false colors in the superimposed image to be presented incorrectly.

[0078] Another design of the method provides that when calculating the superimposed image, the different color channel values of the white-light image are increased and / or decreased position-resolvedly - in particular pixel by pixel - directly or indirectly according to the local intensity values of the spectral image such that the local total brightness is at least approximately maintained, in particular with a relative brightness loss of less than 25%, preferably less than 10%. The local total brightness can in particular be obtained from the sum of the color channel values of all color channels of the white-light image. But this does not necessarily have to be the case. With this solution, it can be achieved that no obvious image brightness of the white-light image is lost in the superimposed image and / or the brightness contrast of the white-light image is still locally substantially maintained despite the superposition in the superimposed image.

[0079] For example, if the white-light image exists as an RGB image with corresponding pixels (each pixel having an R, G, and B color channel value), then the above conditions are met if it is ensured that even after the color channel values of the white-light image are appropriately adapted within the range of superimposed image calculation, the change in the sum of the three signal values R / G / B is less than 20%, preferably less than 10%. However, it should also be noted here that depending on the image format, the image brightness does not necessarily have to be obtained as a simple addition of the color channel values, but there can also be more complex relationships between these parameters, but then the relationships can be considered accordingly.

[0080] Another variant of the method provides that a local intensity threshold I0 is defined for the spectral image, and that the reduction according to the invention of at least one color channel value of the white light image is only carried out locally when the intensity threshold I0 in the spectral image is exceeded at this location (x, y), i.e. when I(x, y)>I0 applies. Here, the intensity threshold I0 should be chosen to be positive. In this case, the following image regions of the white light image can be superimposed on the spectral image with unchanged local color values, in which the spectral image only shows intensity values below the threshold I0. Thereby, the color authenticity of the WLI background can be maintained in image regions that are less relevant with respect to the spectral image. For example, the color channel values to be reduced can be the complementary color values mentioned.

[0081] In other words, it can thus be provided that if the local intensity value I(x, y)=0 or <I0, the reduction of the at least one color channel value (such as the B value) of the white light image can be omitted.

[0082] To solve this task, as mentioned, an image recording device is also proposed, which is designed in particular in the form of a medical visualization system and / or can include an endoscope, an exoscope or a microscope. To solve this task, the device already described at the beginning also includes an image signal processing unit, which is set up to carry out the method as described above and / or as claimed in one of the method claims. Here, the image signal processing unit calculates a superimposed image from the white light image and the spectral image / fluorescence image, in particular taking into account the false color representation of the previously calculated spectral image / false color spectral image.

[0083] The image recording device can also include (in a manner known per se) an excitation light source for generating excitation light. Spontaneous emission and thus fluorescence can be generated with the excitation light, which is detected sensorially by the at least one image sensor when recording the fluorescence image.

[0084] In particular, the image recording device can for example include two image sensors, such as a color image sensor for sensorially detecting the white light image and a second - in particular monochromatic - image sensor, which is set up to detect the spectral image / fluorescence image spatially separated.

[0085] Furthermore, the image recording device can be designed and / or set up analogously as described above with reference to alternative solutions i) to iii) (see also claim 11 for this).

[0086] The present invention will now be described in more detail by way of examples, but is not limited to these examples. Other configurations of the present invention can be obtained from the following description of the preferred embodiments in conjunction with the general description, claims and drawings. In the following description of different preferred embodiments of the present invention, even if the appearance or shape is different, elements that are functionally consistent are given consistent reference numerals. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Shows:

[0088] FIG. 1 shows a scheme in the prior art for spatially separating the detection of white light images and fluorescence images;

[0089] FIG. 2 shows another previously known scheme in the prior art for temporally separating the detection of white light images and fluorescence images;

[0090] FIG. 3 shows a previously known concept in the prior art, namely so-called alpha blending, for superimposing spectral images with white light images;

[0091] Figure 4 Shows a possible implementation of the method according to the present invention at the signal processing level;

[0092] FIG. 5 shows a previously known problem that may occur when superimposing fluorescence images with white light images;

[0093] Figure 6 Shows the application of the method according to the present invention for obtaining a high-quality superimposed image;

[0094] Figure 7 Shows a mapping function by which the intensity values of the spectral image can be mapped into a false color space;

[0095] Figure 8 Shows a superimposing function by which the values of multiple individual color channels of the white light image can be determined for how to adapt when superimposed with the spectral image;

[0096] Figure 9 Shows another superimposing function for the case where the spectral image should be shown in multiple different false colors;

[0097] Figure 10 Shows the application of the method according to the present invention to the superimposition of a spectral image with a white light image, where the spectral image is visualized only in a single unique green false color;

[0098] Figure 11 Shows the HSV color space;

[0099] Figure 12Shows another view of the HSV color space;

[0100] Figure 13 Illustrates, based on a schematic example of the indicated intensity distribution of a spectral image, how different image regions of a spectral image can be adapted according to the invention (individually for each color channel);

[0101] Figure 14 Shows a method according to the invention at the level of image signal processing for the case of a false - color spectral image that is visualized only in a single unique green false color; and finally

[0102] Figure 15 Shows the application of the method according to the invention at the level of signal processing for the case of a false - color spectral image that is to be visualized as an RGB image with multiple different false colors in an overlay image. Detailed description

[0103] Figure 1 shows an image recording device 5, which includes a camera 15 as part of an endoscope. The camera has imaging optics 17 that direct light emitted from an object 18 onto a beam splitter 16. The observed scene is irradiated here with a broadband white - light illumination device with wavelengths in the ultraviolet range and also in the visible range. The object 18 also includes fluorophores that are excited by ultraviolet wavelengths to emit spontaneously in the infrared wavelength range (IR). Of course, fluorophores that can be excited with wavelengths in the near - infrared range can also be used.

[0104] The beam splitter 16 directs infrared wavelengths (i.e., fluorescence) onto a monochromatic image sensor 6b, while visible wavelengths reach a common color image sensor 6a. With the image recording device 5, it is thus possible to sense and detect a white - light image 2 with the image sensor 6a, and simultaneously, but spatially separated therefrom, to sense and detect a spectral image 3 in the form of a fluorescence image 4 with a second image sensor 6b. The image recording device 5 also includes an image signal processing unit 7, which can consist of a plurality of electronic components and is set up to calculate a digital overlay image 1 from the recorded white - light image 2 and the recorded spectral image 3. Then, the overlay image 1 can be displayed on a display unit 13.

[0105] When the image signal processing unit 7 is set up to implement the method of the invention as indicated by the star symbol in Figure 1, the image recording device 5, which is itself previously known from the prior art according to Figure 1, can be designed according to the invention.

[0106] The same applies to the image recording device 5 which is also already known from the prior art per se in FIG. 2. The image recording device differs from such an image recording device in FIG. 1 only in that only a single image sensor 6 is used in FIG. 2, but this image sensor detects the white light image 2 and the fluorescence image 4 separately from each other in time. For this purpose, an alternating illumination is used which consists of the illumination light of the illumination light source 11 in the visible light range and the ultraviolet excitation light of the excitation light source 12, wherein the two light sources 11, 12 illuminate the object 18 alternately in time. Accordingly, the illumination light 19 and the fluorescence 20 respectively arrive alternately at the imaging optics 17 of the image recording device 5.

[0107] FIG. 3 shows a first method known from the prior art, the so-called alpha blending, by means of which the white light image 2 and the spectral image 3 (for example in the form of a false color spectral image 3b) can be superimposed to form a superimposed image 1 by weighted addition on the signal processing level. For this purpose, the white light image 2 is weighted with a coefficient α1 < 1, while the spectral image 3b is globally weighted with the complementary weighting coefficient α2 = 1 - α1. Although this method can in principle avoid additional clipping in the superimposed image 1, the disadvantage of this method is mainly that the image brightness related to the white light image part is lost, as can be seen by comparing the details in the upper right corner of FIG. 3 with the original white light image 2.

[0108] Furthermore, the original / clipped 9 is seen in the white light image 2 of FIG. 3, i.e., the image area presented as white, because all three color channels RGB of the white light image 2 already show full amplitude there. However, as can be seen according to detail A, additional clipping 10, i.e., a white image area generated only by the superposition of the two images 2, 3, may also occur here when the weighting coefficient α1 is not properly selected. And in detail B, the reduced image brightness 22 caused by the application of alpha blending is seen. Due to the global reduction of the image brightness of the white light image 2, it can also be seen in detail C that the original optical reflection 23 still visible in the original white light image 2 can no longer be seen in the superimposed image 1. The result is thus the following presentation in the superimposed image 1, which provides a worse spatial impression of the observed scene compared to the original white light image 2.

[0109] To understand the method according to the invention for generating a high-quality superimposed image 1, first Figure 13 it is helpful: The figure shows at the top a schematic diagram of the intensity distribution I(x, y) of the spectral image 3a. As can be seen according to the scale, the intensity of spectrally narrow-limited light (for example fluorescence) is detected / visualized position-resolved in the spectral image 3 with an 8-bit resolution (i.e., in the value range of 0..255).

[0110] With the aid of the mapping function MF(I), a false-color representation FF(x, y) of the spectral image 3a in the form of a false-color spectral image 3b can be generated from the original spectral image 3a. As Figure 13 can be seen, in the false-color spectral image 3b here, different intensity levels are visualized in five different false colors, namely from black via blue, cyan, green to yellow. Thus, the yellow image region Y in the false-color spectral image 3b corresponds to the image region with the highest intensity in the original spectral image 3a.

[0111] Thus, there are different local false colors FF(x, y) in the false-color spectral image 3b, for example a blue false color B and a yellow false color Y. If the white-light image 2 is recorded, for example, as an RGB image with three color channels R / G / B, then each local false color in the local false colors of the false-color spectral image 3b can be assigned a corresponding complementary color channel. For example, the blue channel B of the white-light image is complementary to the yellow false color Y. With respect to the corresponding complementary color channel value B of the white-light image 2 that is complementary to the yellow local false color Y, the method according to the invention now provides that this color channel value B is locally and selectively reduced compared to the remaining two local color channel values R and G of the white-light image 2. However, this reduction does not occur globally as in alpha blending, but occurs indirectly in relation to the relevant local intensity value I(x, y) of the spectral image 3a: For example, for the square image region in the center of the spectral image 3a (which has a high local intensity I and which is visualized in the false-color spectral image 3b with the yellow false color Y), it is advantageous to specifically reduce the local color channel value B there in the corresponding image region of the white-light image 2. This can be seen at the "-" sign in Fig. (e), which shows the component / correction function ΔB(x, y) of the superimposition function OF(I), with the aid of which the white-light image 2 and the spectral image 3 are specifically superimposed. Because as can be seen in the schematic diagram (e): The complementary color channel value B in the central square image region is specifically reduced, and the central square image region corresponds to the central region of the original spectral image 3 shown in the false color Y.

[0112] Pair Figure 13 Observation of Figs. (c) and (d) here shows that the remaining color channel values R and G of the white-light image 2 in the central square image region (which is visualized in the false-color spectral image 3b with the yellow false color Y) are even emphasized. Because red and green are not exactly complementary to the false color Y to be visualized at this location (x, y).

[0113] Figure 5 shows another example of the prior art of the superposition of a white light image 2 and a spectral image 3 in the form of a fluorescence image 4, which spectral image should be visualized as a false-color spectral image 3b by means of false-color rendering. The image 3b here includes false colors: black, blue, and yellow. Here, the black image area corresponds to the low-intensity area of fluorescence, blue corresponds to the medium-intensity area of fluorescence, and yellow corresponds to the high-intensity area of fluorescence. In the case of the superposition 27 common in the prior art of the two images 2, 3b, it can occur that in the image area (e.g., see detail B) of the superposition image 1 where the false color (here yellow) is to be visualized, an additional clipping 10 appears, which then causes these image areas to no longer be visualized in the correct false color but in white, as can be seen in detail D of the superposition image 1. Thus, at this location, a color shift 28b of the yellow false color to be presented at this location occurs.

[0114] However, in addition, in the superposition image 1 of FIG. 5, another color shift 28a is also seen in the following image area where the actual blue false color should be visualized. But in the relevant white light image 2, the image area there is colored red. Thus, in the case of adding the blue false color and the red color in the white light image 2, a purple color is produced, but the purple color is not included in the used false-color scale from black via blue to yellow. Accordingly, it will be difficult for the surgeon here to assign a fluorescence intensity to this image area of the superposition image 1 according to the false-color scale.

[0115] Figure 6 Now it is shown how the problem of color shift of false colors can be avoided by means of the solution according to the invention: In the left figure, the white light image 2 is seen, which white image shows the tissue surface with blood vessels, and the blood vessels are visualized in red. Thus, the white light image 2 uses true colors and gives the surgeon a realistic impression of the observed surgical scene. In addition, as a color bar, a false-color scale is also seen, which continuously ranges from black via blue, cyan, dark green, light green to yellow, and which false-color scale can (comparable to the false-color spectral image 3b of FIG. 5) be used to visualize the intensity of the spectral image 3, in particular the intensity of the fluorescence image 4. The superposition image 1 produced by means of the method according to the invention (this superposition image is in Figure 6As can be seen in the right part of ), on the one hand, the original white light image 2 is now reproduced as a colored background image, but at the same time a false color representation of the fluorescence signal is superimposed on it according to the original spectral image 3. Unlike the example also in FIG. 5, the relevant color shift 28 no longer occurs; instead, as shown in detail D, image areas with high fluorescence signal intensity are visualized in the correct yellow false color. This is achieved in that the blue color channel value B of the white light image 2 on the left is specifically reduced in these image areas. At the same time, in these image areas, the non-complementary red and green color channel values R and G are slightly highlighted, so that despite the reduction in the blue color channel value, the brightness of the image point in the white light image 2 is essentially maintained. As a result, in particular the local brightness changes and therefore the contrast of the white light image 2 in the superimposed image 1 can be maintained. In addition, unlike the alpha blending according to FIG. 3, no global brightness loss is observed in the superimposed image 1.

[0116] Figure 7 An example of a possible mapping function MF(I) is shown, with which different intensity values I(x, y) of the original spectral image 3 can be visualized with the aid of the green false color G. The mapping function thus assigns each intensity value an associated color channel value G(I) on a numerical scale of 0...255.

[0117] Figure 10 It is shown how a false-color spectral image 3 b can be obtained in this way by visualizing the corresponding original intensities I(x, y) of the original spectral image 3 with the aid of only a single green false color. In this example, it can be seen that when the method according to the invention is applied, no noticeable color shift is observed in the superimposed image 1 in image regions where there is a noticeable intensity in the spectral image 3. However, it can still happen, as shown in detail A, that a certain color shift of image signal portions of the original white-light image 2 can be observed in other image regions. However, this is not critical for the application as long as the surgeon can still see an effective false-color presentation of the spectral image 3 in the superimposed image 1 as before.

[0118] Figure 11 A possible color space is shown for this purpose, namely the HSV color space, in which false color representations can be encoded. Figure 10 In the example of , it is conceivable that the only green pseudocolor is retained in terms of its color value H, but either the color brightness 33 (radial axis) and / or the color saturation S32 (vertical axis) are changed, respectively, so that different intensity levels can be visualized with the green pseudocolor. Figure 11 The mapping function MF is involved a As shown by the thin black arrow in (I), Figure 10In the false-color rendering, the H value of the green false color remains constant and only the color brightness 33 changes (changing in the radial direction in the color space).

[0119] In contrast, in Figure 6 the example, different false colors from black to yellow are used to visualize the spectral image 3. Looking at Figure 12 (This figure shows a top view of the HSV color space), this can be understood as defining a continuous path in the color space in this scheme, where different color values H are assigned according to the intensity values. Accordingly, the mapping function MF(I) can include multiple components, such as R(I), B(I), and G(I). Therefore, in Figure 6 the example, the relevant mapping function will first increase the blue color channel value B in the case of low intensity values, and then decrease the blue color channel value B in the case of higher intensity values. Since the yellow color is visualized by the red and green color channel values, the mapping function assigns relatively high color channel values in the green and red color channels of the false-color spectral image 3b to high signal values.

[0120] In the method according to the invention, the superimposed image 1 is preferably calculated with the aid of the superimposition function OF(x, y). This superimposition function thus determines for each pixel of the superimposed image 1 in a position-resolved manner by which amount the corresponding local color channel value of the white-light image 2 should be increased or enhanced, where the superimposition function OF(x, y) indirectly takes into account the corresponding local intensity value I(x, y) of the original spectral image at least. As shown in Figure 7 in the mapping function (the mapping function can produce a false-color spectral image 3b as shown in Figure 10 ), the superimposition function OF(I(x, y)) can look, for example, as shown in Figure 8 . For low intensity values, the color channel values R / G / B of the white-light image 2 actually remain unchanged because the corresponding amounts ΔG / ΔR / ΔB by which the corresponding color channel values are increased or decreased are very small for low intensity I. At high intensities, for example, in the case of a value of 250, the amount ΔG(I) by which the green color channel value G of the white-light image 2 should be increased is correspondingly high, and then the remaining two color channel values R and B, which are complementary to the green false color to be visualized, are exactly decreased, which can be understood from the large negative values of the coefficients ΔR and ΔB for high intensity values I in the figure of Figure 8 .

[0121] In contrast, Figure 9 shows for Figure 6Superposition function for the case of an application example, where the spectral image 3 is visualized with the shown continuous false-color scale from dark blue to yellow. At low intensity values, the blue false color should be visualized here, while at high intensity values the yellow false color should be visualized. Accordingly, Figure 9 the superposition function indicates at low intensity values of I < 100 that the blue color channel value B of the white light image 2 increases accordingly (ΔB > 0). While at high intensity values, the yellow false color should be visualized according to the false-color scale. Accordingly, in that case the complementary blue color channel value B of the white light image 2 is locally reduced respectively, i.e., reduced according to the local intensity value I(x, y) of the original spectral image 3, as can be seen from the negative value of the parameter ΔB for high intensity values I in Figure 9 . While the remaining color channel values G and R of the white light image 2 increase at high intensity values; the values of ΔG and ΔR in the Figure 9 figure are correspondingly high there.

[0122] As Figure 13 shown, a possible implementation of the mapping from the original intensity distribution I(x, y) of the original spectral image 3a to the false-color rendering FF(x, y) in the false-color spectral image 3b can be understood according to Figure 4 : It can be seen at the lower left image edge how the original spectral image 3a is converted into the false-color spectral image 3b with the aid of the mapping function MF(I(x, y)) - in the case of visualization with the three color channel values RGB. The method according to the invention can now be implemented on the signal processing level in the following way, namely, according to the respective local color channel values (R Figure 4 / G FF / B FF / B FF ) of the false-color spectral image 3b (all of these local color channel values are directly related to the corresponding original local intensity value I(x, y) of the original spectral image 3a based on the mapping function) with different degrees of weight, i.e., according to the weighting factors (b i , r i and g i ) are settled with the color channel values (R / G / B) of the white light image 2 in order to ultimately obtain the corresponding color channel values R, G, B of the superimposed image 1 thereby. Thus, this weighting / settling is carried out pixel by pixel at each location x, y of the superimposed image, where the respective weight in each pixel can be changed according to the respective color signal values R FF , G FF and B FF of the false-color spectral image 3b, i.e., indirectly according to the relevant original intensity value I(x, y) of the original spectral image 3a. Figure 13The following three figures (c), (d) and (e) exemplarily show such a spatial variation of the weights for the three color channels R / G / B of the white light image 2.

[0123] Figure 14 shows the application of the method according to the invention to Figure 10 the case of, i.e., only a unique green false color G FF of the spectral image 3: Here, the false color spectral image 3b only contains a unique green color value G FF , which green color value can be locally determined, for example, with the aid of Figure 7 the mapping function directly according to the corresponding local intensity value I(x, y) respectively. For this green false color, the green color channel value G of the white light image 2 is non-complementary, while the red color channel R and the blue color channel B of the white light image 2 are exactly complementary. Here, in the RGB color space, the increase or decrease of the corresponding complementary color components corresponds to a change in the saturation of the superimposed color or a color shift in the direction of the saturated superimposed color, and the superimposed color is used to present the fluorescence signal in the form of an image overlay.

[0124] Correspondingly, as can be seen in Figure 14 , with the weighting factors g1 < 0, g2 > 0 and g3 < 0, the two complementary color channel values R and B of the white light image 2 are reduced position-resolved and selectively relative to the non-complementary color channel value G (but only if a distinct color value G FF is to be visualized at this location), while if a green false color is to be visualized in the overlay image 1 at this location, the non-complementary color channel value G of the white light image 2 is emphasized. Thus, by this measure, it can be ensured that the green false color in the overlay image 1 is visualized without a relevant color shift, more precisely also for the following image points of the white light image 2, which would otherwise show the complementary mixed color (e.g., orange or purple hues).

[0125] As can also be seen in Figure 14 , here, in order to generate the overlay image 1, pixel by pixel, the corresponding ratio (the corresponding local color channel values R / G / B of the white light image 2 and the relevant local signal value G of the spectral image 3b FF should be settled at this ratio) is selected position-dependently, i.e., indirectly according to the original local intensity value I(x, y) of the corresponding pixel (x, y) of the original spectral image 3a. Thereby, the corresponding weights between the local color channel values R / G / B of the white light image 2 and the corresponding local signal value G of the spectral image 3b FF can be changed position-dependently according to the original local intensity value of the spectral image 3a.

[0126] In contrast,Figure 15 shows a possible implementation of the method according to the invention at the signal processing level for the case of Figure 6 , namely, the visualization of the spectral image 3 by means of a continuous false-color scale ranging from dark blue to yellow: Here, the false-color spectral image 3b (which can be obtained from the original spectral image 3a by means of a mapping function) can now include three different color channel values R, G, B, the individual values of which vary according to the local intensity value I(x, y). By means of the weighting coefficient a i , a certain color transformation can first be applied to the white-light image 2. The same applies to the false-color spectral image 3b, which can be color-transformed by means of the weighting coefficient s i . By means of a further weighting coefficient b i , r i and g i , the color channel values R / G / B of the false-color spectral image 3b can now be settled with the color channel values R / G / B of the white-light image 2 in order to thereby calculate the superimposed image 1.

[0127] It can also be seen here again that, for example, if the false-color spectral image 3b should locally visualize high signal values in the red false color R, the ways implemented by the weighting coefficients r1>0, r2<0, and r3<0 act on the three color channel values R / G / B of the white-light image 2 accordingly. Thereby, the red channel value R of the white-light image 2 is correspondingly emphasized, while the remaining color channel values G and B of the white-light image 2 are correspondingly attenuated. Therefore, by Figure 15 the scheme shown in can achieve the superimposition of the two images 2 and 3, where the original false-color rendering of the spectral image 3b can be reproduced almost unchanged in the superimposed image 1, which ensures for the surgeon observing the superimposed image 1 that the surgeon can reliably infer the original intensity I(x, y) of the original spectral image 3a based on the color of the superimposed image 1.

[0128] The totality of the weighting coefficients ri, bi, gi associated with the intensity-dependent color channel values R(I) / G(I) / B(I) of the false-color spectral image 3b in Figure 15 can thus be understood as a superimposition function OF(x, y) in the sense of the present invention, because it is stipulated by the totality of the weighting coefficients: by which amount the corresponding local color channel values R(x, y), G(x, y), B(x, y) of the white-light image 2 should be increased or decreased when calculating the superimposed image 1, namely according to the sign of the corresponding weighting coefficient and according to the color channel values R(I) / G(I) / B(I) of the false-color spectral image 3b. Therefore, in the Figure 15 example, the weighting coefficients r i , b i , gi It is determined separately which color channel values of the white light image 2 are to be attenuated as complementary color channel values with respect to a specific false color (defined by the three color channel values R, G, B of the false color spectral image 3b), and which of these color channel values of the white light image 2 are to be enhanced exactly as non-complementary color channel values.

[0129] In summary, in order to generate a realistic true-color superimposed image 1 related to at least one false color FF(x, y) used for visualizing the spectral images 3a, 3b, a method and a related image recording device 5 are proposed. The image recording device can achieve avoiding an undesired color shift of such false colors in the superimposed image 1, and basically maintaining the high image brightness of the white light image 2 shown as the image background and the original image contrast of the white light image 2 in the superimposed image 1 despite the superimposition. For this purpose, it is provided according to the present invention that at least one color channel value of the white light image 2 that is complementary to the false color to be presented at the corresponding local part is locally and specifically reduced, more precisely, the higher the intensity I(x, y) of the spectral image 3a that is also to be visualized in the superimposed image 1 at this part (x, y) (see Figure 6 ), the more it is reduced.

[0130] List of reference numerals

[0131] 1 Superimposed image

[0132] 2 White light image (WLI)

[0133] 3 Spectral image (SI)

[0134] 4 Fluorescence image (FI)

[0135] 5 Image recording device

[0136] 6 Image sensor

[0137] 7 Image signal processing unit

[0138] 8 Visualization system

[0139] 9 Original cropped-image region ("original crop" in 2)

[0140] 10 Additional cropped-image region ("additional crop" in 1)

[0141] 11 Irradiation light source

[0142] 12 Excitation light source

[0143] 13 Display unit

[0144] 14 Image signal (of 6)

[0145] 15 Video camera

[0146] 16 beam splitter

[0147] 17 imaging optical device

[0148] 18 object

[0149] 19 irradiation light

[0150] 20 fluorescence

[0151] 21 false color

[0152] 22 reduced loss of image brightness / image contrast

[0153] 23 optical reflection (on wet tissue)

[0154] 24 false color calculation unit

[0155] 25 color channel values (e.g., with a numerical range from 0 to 255 in 8-bit resolution; e.g., 2 / 3 / 1)

[0156] 26 presentable / usable numerical range (for 25)

[0157] 27 superposition / overlay (of 2 and 3)

[0158] 28 (false) color shift (color drift)

[0159] 29 false color scale

[0160] 30 numerical range for intensity I(x, y) (for 3)

[0161] 31 color value (hue, H)

[0162] 32 color saturation (saturation, S)

[0163] 33 color brightness (value, V)

[0164] 34 presentable color space

[0165] 35 false color mapping

Claims

1. A method for generating a superimposed image (1), - Among them, wherein the superimposed image (1) visualizes the image signal part of a white light image (2) (WLI) and the image signal part of a spectral image (3a), in particular a fluorescence image (4), together, - wherein the local intensity values I(x, y) of the spectral image (3a) are visualized in the superimposed image (1) by means of at least one - in particular corresponding - local false color FF(x, y), - wherein the white light image (2) is visualized in the superimposed image (1) by means of grayscale or by means of true color, and - wherein, in order to generate the superimposed image (1), the local color channel values (R / G / B) of the white light image (2) are intersettled with the associated local signal values (R / G / B or GW) of the spectral images (3a, 3b), characterized in that - at least one complementary color channel value (B) that is complementary to the white light image (2), in particular complementary to the corresponding local false color FF(x i , y i ), is locally and selectively reduced relative to the remaining local color channel values (R, G) of the white light image (2), more precisely not globally, but directly or at least indirectly in accordance with the relevant local intensity value I(x, y) of the spectral image (3a). - preferably, the remaining color channel values (R, G) of the white light image (2) are locally and selectively maintained or even increased directly or indirectly as a function of the local intensity values I(x, y) of the spectral image (3a), and / or - in particular, the higher the corresponding local intensity value I(x, y) of the spectral image (3a), the more strongly the at least one complementary color channel value (B) of the white light image (2) is reduced as a function of position, and / or - in particular, as a function of the associated local intensity value I(x, y) of the spectral image (3a), the at least one complementary color channel value (B) of the white light image (2) is reduced to different extents, in particular not at all, at different positions (x, y) in the superimposed image (1).

2. A method for generating a superimposed image (1), in particular according to the preamble of claim 1 or according to the method according to claim 1, - Among them, wherein the superimposed image (1) visualizes the image signal part of a white light image (2) (WLI) and the image signal part of a spectral image (3a), in particular a fluorescence image (4), together, characterized in that - in order to generate the superimposed image (1), preferably pixel by pixel, the respective ratio used for intersettling the respective local color channel values (R / G / B) of the white light image (2) with the associated respective local signal values (R / G / B) of the spectral images (3a, 3b) is selected as a function of position and directly or indirectly as a function of the respective original local intensity value I(x, y) of the respective pixel (x, y) of the spectral image (3a), - in particular such that the respective weights between the local color channel values (R / G / B) of the white light image (2) and the local signal values of the spectral images (3a, 3b) change as a function of position, as a function of the original local intensity value I(x, y) of the spectral image (3a) - in particular pixel by pixel.

3. A method for generating a superimposed image (1), in particular according to the preamble of claim 1 or according to claim 1 and / or according to claim 2, - Among them, The superimposed image (1) visualizes the image signal portion of the white light image (2) (WLI) and the image signal portion of the spectral image (3a), in particular the fluorescence image (4), together. - In order to generate the superimposed image (1), the respective ratio used for calculating the mutual settlement of the local color channel values (R / G / B) of the white light image (2) and the associated local signal values (R / G / B) of the spectral images (3a, 3b) is selected position-dependently - preferably pixel by pixel. It is characterized in that - The respective ratio is calculated using a superimposition function OF(I) = f(I(x, y)) (superimposition function), which is defined directly or indirectly based on the respective local intensity values I(x, y) of the original spectral image (3a): by which amount the respective local color channel values of the white light image (2) are increased / enhanced / emphasized or decreased / diminished / reduced when calculating for generating the superimposed image (1). - In particular, the superimposition function OF(I) determines which color channel values (R / G / B) of the white light image (2) are attenuated as complementary color channel values (B) with respect to a specific local false color FF(x i , y i ) in the superimposed image (1), and / or which color channel values among these color channel values (R / G / B) of the white light image (2) are enhanced as non-complementary color channel values (R, G). -Preferably so that the local false color FF(x i , y i ) to be presented is also presented in true color / without significant color shift in the superimposed image (1).

4. The method according to any one of the preceding claims 1 to 3, wherein, In particular, as an intermediate step when calculating the superimposed image (1), a false color representation FF(x, y) of the original spectral image (3a) in the form of a false color spectral image (3b) is generated from the original spectral image (3a) by means of a mapping function MF(I), and the false color spectral image is at least partially visualized in the superimposed image (1). - Preferably, the mapping function MF(I) assigns a specific local false color FF(H,S,V) to the respective local intensity values I(x, y) of the original spectral image (3a), and / or - In particular, the mapping function MF(I) maps into a false color subspace. In particular, the false color subspace includes: - Only a single unique false color value H, but for example includes different color saturation values S and / or different color brightness values H, or - Multiple different false color values H, preferably the multiple different false color values form a continuous path within the false color subspace.

5. The method according to claim 4, wherein, The false color spectral image / the false color representation FF(x, y) visualizes / can visualize false colors: blue, cyan, green, and yellow, and / or - The mapping function MF(x, y) describes a continuous false color change from blue to cyan to green to yellow in the false color subspace, and / or - Particularly preferably, the high intensity values I(x, y) of the spectral image (3) are assigned to the yellow false color, and the low intensity values I(x, y) of the spectral image (3) are assigned to the blue false color. - In particular, the false color spectral image FF(x, y) also includes black as a false color.

6. The method according to any one of the preceding claims, wherein, The false color spectral image / the false color representation FF(x, y) visualizes / can visualize the false color green, and preferably the false color black, and / or - The mapping function MF(x, y) describes a continuous brightness change and / or saturation change in the false color subspace for at least one false color, in particular for the false color change from black to green. - Preferably, the high intensity value I(x, y) of the spectral image (3) is assigned to the green false color, and the low intensity value I(x, y) of the spectral image (3) is assigned to the black false color.

7. The method according to any one of the preceding claims, wherein - Especially in the mentioned location-related settlement, according to the local original intensity value I(x,y) of the spectral image (3a), the local color channel values (R, G, B) of the white light image (2) and the corresponding false color signal values (R, G, B) of the false color spectral image (3b) are mutually settled with different location weights respectively.

8. The method according to any one of claims 4 to 7, wherein -The corresponding local false color FF(x i , y i ) of the false color spectral image (3b) defines a corresponding complementary color KF(x, y), and at least one color channel value of the white light image (2) corresponding to the complementary color KF(x, y) is reduced position-dependently - in particular pixel by pixel - - especially in order to thereby limit a corresponding local false-color FF(x i , y i ) color shift (28) (color drift) in the resulting superimposed image (1).

9. The method according to any one of the preceding claims, wherein, The superimposed image (1) is calculated based on a superimposing function OF(x, y) (superimposing function), especially a predefined, preferably user-adaptable superimposing function, which directly or at least indirectly prescribes according to the corresponding local intensity value I(x, y) of the original spectral image (3a): by which amount the corresponding local color channel values (R / G / B) of the white light image (2) are increased / enhanced / emphasized or decreased / weakened / reduced. - Preferably, the local color channel values of the white light image (2) are increased / enhanced / emphasized when the local color channel values of the white light image correspond to the relevant local false color FF(xy), and the local false color is visualized / should be visualized as an image signal part of the spectral image (3a) at the location (x, y) in the superimposed image (1), and / or - Preferably, the local color channel values of the white light image (2) are decreased / weakened / reduced when the local color channel values of the white light image are complementary to the relevant local false color FF(xy), and the local false color is visualized / should be visualized as an image signal part of the spectral image (3a) at the location (x, y) in the superimposed image (1).

10. The method according to any one of the preceding claims, wherein, Both the white light image (2) and the spectral image (3) / fluorescence image (4) are sensed and detected by a single image recording device (5), and the image recording device is especially in the form of an endoscope, an external endoscope or a microscope and / or is designed according to claim 13.

11. The method according to any one of the preceding claims, wherein, The white light image (2) and the spectral image (3) / the fluorescence image (4): i) Are sensed and detected separately from each other spatially. Especially, - By means of at least two separate image sensors (6a, 6b), and / or - By means of different color filters at the pixel level of the image sensor, especially an RGBX sensor or a hyperspectral sensor, or ii) Are sensed and detected separately from each other temporally. Especially - By means of temporally varying illumination, and / or - By alternately detecting two images (2, 3 / 4) over time using the image sensor (6), or iii) Are separated only by intelligent signal processing. Especially, in this case, the two images (2, 3 / 4) are sensed and detected simultaneously and / or by only a single image sensor.

12. The method according to any one of the preceding claims, wherein, The portion of the spectral image (3) in the superimposed image (1) is visualized in false color as FF(x, y), and / or - the portion of the white light image (3) in the superimposed image (1) - is visualized as a colored white light background image, preferably in true color, or - is visualized as a monochromatic gray-scale background image.

13. The method according to any one of the preceding claims, wherein, The white light image (2) is visualized as a gray-scale image, or - the white light image (2) includes color information of at least two different color channels, and / or - the original spectral image (3) visualizes the intensity values I(x, y) sensed and detected of spectral light signals - especially fluorescence signals.

14. The method according to any one of the preceding claims, wherein, When calculating the superimposed image (1), it is checked position-resolved - especially pixel by pixel - whether, by superimposing the image signal portion of the white light image (2) with the image signal portion of the spectral image (3a), the available and / or maximally presentable value range (26) in the superimposed image (1) is exceeded, and if this is the case, the image signal portion of the white light image (2), especially the at least one complementary color channel value (B) of the white light image (2), is reduced to such an extent that the value range (25) is locally maintained. - especially such that additional cropping (10) in the superimposed image (1) is avoided, - preferably, the original cropped-image region (9) of the white light image (2) that already has an excess of the presentable value range, especially in the form of white image spots, continues to be presented in the superimposed image (1), but especially in the correct false color, as long as the spectral image (3) has a distinct intensity I(x,y) corresponding to the respective cropped-image region (9).

15. The method according to any one of the preceding claims, wherein, When calculating the superimposed image (1), the different color channel values (R / G / B) of the white light image (2) are increased and / or decreased position-resolved - especially pixel by pixel - directly or indirectly according to the local intensity values I(x,y) of the spectral image (3a) such that the local total brightness - especially the local total brightness formed by the sum of the color channel values (R / G / B) of all color channels of the white light image (2) - is at least approximately maintained, especially with a relative brightness loss of less than 25%, preferably less than 10%. - especially such that no distinct image brightness of the white light image (2) is lost in the superimposed image (1), and / or the brightness contrast of the white light image (2) can still be locally substantially maintained despite the superimposition in the superimposed image (1).

16. The method according to any one of the preceding claims, wherein, A local intensity threshold I0 is defined, and only when the intensity threshold I0 in the spectral image (3) is exceeded at the location (x, y), i.e., when I(x i , y i ) > I0, is the reduction of at least one color channel value (B) of the white light image (2) performed locally.

17. An image recording device (5), especially an image recording device in the form of a medical visualization system (8), which is designed, for example, as an endoscope, an exoscope or a microscope, the image recording device having: - at least one image sensor (6), which is set up for the sensed detection of a white light image (2) and a spectral image (3), especially a fluorescence image (4), and The image recording device has: - An image signal processing unit (7), which is configured to calculate a superimposed image (1) from a white light image (2) recorded by using the at least one image sensor (6) and a spectral image (3), in particular a fluorescence image (4), recorded by using the at least one image sensor (6). - Preferably output in the form of a digital superimposed image (1) (superimposition function). It is characterized in that - The image signal processing unit (7) is configured to perform the method according to any one of the preceding claims when the image signal processing unit (7) calculates the superimposed image (1).

Images