Image rendering method and electric equipment

Abstract

The invention discloses an image rendering method and electric equipment, and the method comprises the steps: obtaining a target image, and enabling the target image to be obtained through the rendering of a to-be-processed image; the image rendering comprises a watercolor toning step, the watercolor toning step comprises toning processing based on a synthetic texture image, and the synthetic texture image is obtained by fusing a plurality of texture images. According to the method, image rendering can be performed on the to-be-processed image to obtain the target image, and toning is performed based on the synthetic texture image during image rendering, so that watercolor toning only needs to be performed once, the rendering time is shortened while the rendering effect is ensured, and the rendering efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of image processing, and in particular to an image rendering method and an electrical device. Background Technology

[0002] In related technologies, when rendering images, the watercolor color grading step uses multiple texture maps as input textures in sequence to perform color conversion on the image. This requires multiple color grading processes, which increases the amount of computation, makes rendering time longer, and results in low rendering efficiency. Summary of the Invention

[0003] This invention aims to at least solve one of the technical problems existing in the prior art. Therefore, one objective of this invention is to propose an image rendering method that can render an image to be processed to obtain a target image. During image rendering, color adjustment is performed based on a synthesized texture map, thus requiring only one watercolor-like color adjustment. This reduces rendering time and improves rendering efficiency while ensuring rendering quality.

[0004] The second objective of this invention is to provide an electrical device.

[0005] To address the aforementioned problems, a first aspect of the present invention provides an image rendering method, comprising: obtaining a target image, wherein the target image is obtained by rendering an image to be processed; the image rendering includes a watercolor color adjustment step, wherein the watercolor color adjustment step includes color adjustment processing based on a synthesized texture map, wherein the synthesized texture map is obtained by fusing multiple texture maps.

[0006] According to the image rendering method of the present invention, in the watercolor color adjustment step, the color adjustment is performed based on the synthesized texture map, that is, only one watercolor color adjustment is needed, which reduces the amount of computation. Moreover, the synthesized texture map is a fusion of multiple texture maps, which can achieve the same rendering effect as when multiple color adjustments are performed based on multiple texture maps separately. Thus, while ensuring the rendering effect, the rendering time is reduced and the rendering efficiency is improved.

[0007] In some embodiments, the synthesized texture map is obtained by fusing multiple different noise texture maps; the multiple different noise texture maps are obtained by simulating different properties of watercolor paint.

[0008] In some embodiments, the composite texture map is obtained by fusing Gaussian noise texture, Berlin noise texture, and paper texture.

[0009] In some embodiments, image rendering further includes image abstraction processing, which includes filtering the acquired image using the Kuwabara filtering algorithm.

[0010] In some embodiments, the Kuwabara filtering algorithm includes obtaining the variance or standard deviation of pixel values ​​in each neighborhood of the target pixel of the image to be processed, wherein the iteration step size used in obtaining the variance or standard deviation is greater than the iteration step size of the standard Kuwabara filtering algorithm.

[0011] In some embodiments, image rendering further includes edge détenting and / or edge dithering.

[0012] In some embodiments, edge jitter processing includes replacing the pixel value of each pixel in the acquired image with the pixel value at an offset position corresponding to the pixel; the offset position is determined based on a random value corresponding to the pixel; the random value is assigned to each pixel in the acquired image based on the synthesized texture map.

[0013] In some embodiments, the images are acquired based on images taken by a drone.

[0014] In some embodiments, the image rendering method is implemented using a graphics processing unit based on the OpenGL framework.

[0015] A second aspect of the present invention provides an electrical device for implementing the image rendering method described in the above embodiments.

[0016] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0017] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a flowchart of an image rendering method according to an embodiment of the present invention; Figure 2 (1) is a schematic diagram of an image rendered according to an embodiment of the present invention, and (2) is a schematic diagram of noise texture used to reconstruct the image by reverse calculation according to an embodiment of the present invention; Figure 3 This is a schematic diagram of an image after Kuwabara filtering according to an embodiment of the present invention; Figure 4 This is a schematic diagram of an image processing procedure taken by a drone according to an embodiment of the present invention; Figure 5 This is a flowchart of an image rendering step according to an embodiment of the present invention; Figure 6 (1) is a schematic diagram of an image before rendering according to an embodiment of the present invention, and (2) is a schematic diagram of an image after watercolor color correction according to an embodiment of the present invention; Figure 7 This is a structural block diagram of an electronic device according to an embodiment of the present invention; Figure 8 This is a structural block diagram of a vehicle according to an embodiment of the present invention.

[0018] Figure label: 200 vehicles; 100 electronic devices; Processor 101; Memory 102; Detailed Implementation The embodiments of the present invention are described in detail below. The embodiments described with reference to the accompanying drawings are exemplary. The embodiments of the present invention are described in detail below.

[0019] Non-realistic photography (NPR) involves processing photographic data to create special artistic effects, such as hand-drawn, watercolor, oil painting, or cartoon styles. This technique has been extensively researched in fields like games and film, with watercolor-style special effects filters being one such branch. As technology has evolved, the methods used in this research have also changed, primarily falling into several technical categories: physically based simulation, brushstroke simulation, texture compositing, and AI (Artificial Intelligence) generation.

[0020] The technology based on physical property simulation is the earliest and also the most complex. For example, existing research abstracts the watercolor painting process into the interaction between water, paper, and paint. This interaction process is modeled, identifying several typical characteristics of watercolor painting and their physical movement patterns. Based on this modeling, a watercolor rendering algorithm is implemented using a computer program to simulate this process. However, this type of algorithm is complex to design, computationally intensive, and requires modeling and implementation at the microscopic particle level. It is difficult to develop high-performance rendering solutions and is not suitable for use in automotive camera scenarios.

[0021] Similar to brushstroke simulation-based solutions, this approach also involves modeling and simulating the drawing process. These algorithms divide the process into two parts: stroke generation and stroke effect processing. Stroke generation requires simulating human drawing habits to generate logically consistent strokes. This process involves too many variables, making objective analysis difficult, and the interaction between multiple strokes is also a significant factor. Furthermore, the simulation of brushstrokes inevitably involves a rendering order, hindering parallel computational optimization. Therefore, this invention did not reference this type of algorithm.

[0022] AI-generated algorithms are currently a mainstream research direction. Through extensive computation and high-parameter model structures, they aim to learn the artistic features of watercolor paintings in greater detail. A unique advantage of these algorithms is their high scalability; they can learn specific artistic styles or the style of particular painters to generate targeted images. However, their disadvantages are also significant. First, they require very high-quality datasets and involve substantial data preprocessing. Second, both training and inference consume enormous computational resources, resulting in high costs. In vehicle-mounted deployments, these algorithms face a dilemma between balancing computational power and accuracy.

[0023] The algorithm based on texture synthesis is the technical foundation of this invention. This type of algorithm mainly consists of two parts: image filtering algorithm and texture overlay algorithm. The image filtering algorithm aims to reduce the dispersion of image colors in the color space, thereby reducing the realism of the image and enhancing its artistic features. Texture overlay further simplifies the modeling, abstracting the physical motion of watercolor into the superposition of several random processes, simulating the uneven settling of pigments by selecting appropriate random processes. Then, the simulation of the watercolor effect is achieved by designing conversion formulas for color concentration and color change. Previous research has made relatively little use of GPUs (Graphics Processing Units), and the related GPU technology stack is not complete. Therefore, most studies have focused on the simulation of effects, and the CPU (Central Processing Unit) processing algorithms used are not efficient for image rendering, leaving room for optimization.

[0024] Current image rendering workflows are divided into two parts: an abstraction step and a watercolorization step. The abstraction step reduces image details and realism; some studies also involve color correction to make the image's color scheme closer to watercolor rather than photographic. The watercolor effect, on the other hand, simulates multiple watercolor features one by one and derives a color transformation formula based on the reflection formula of pigment concentration under light, establishing a texture-density-color conversion relationship. Existing research mostly focuses on adjusting the above workflow and optimizing the effects, with relatively little research on reducing computational cost.

[0025] To address the above problems, the first aspect of this invention provides an image rendering method. This method can render an image to be processed to obtain a target image. During image rendering, color adjustment is performed based on a synthetic texture map, thus requiring only one watercolor color adjustment. This reduces rendering time and improves rendering efficiency while ensuring rendering effect.

[0026] The following is for reference. Figure 1 A method for rendering an image according to an embodiment of the first aspect of the present invention is described, such as... Figure 1 As shown, the method includes at least step S01.

[0027] Step S01: Obtain the target image, which is obtained by rendering the image to be processed.

[0028] The image rendering process includes a watercolor color grading step, which involves color grading based on a composite texture map, which is obtained by fusing multiple texture maps.

[0029] Specifically, the image rendering method proposed in this invention is used for watercolor-style special effects filters in non-photorealistic photography techniques. That is, it can convert ordinary photos or digital images into images with a watercolor painting effect. The image to be processed can be understood as an unrendered image, and the target image can be understood as a rendered image. During image rendering, the image to be processed undergoes watercolor-style color correction, and a synthetic texture map is used for color adjustment to obtain the target image. This makes the target image a rendered image with a watercolor painting effect. The synthetic texture map is obtained by fusing multiple texture maps; therefore, only one watercolor-style color correction is needed to obtain the target image.

[0030] For example, this invention is the first to use texture maps to simulate the material characteristics of painting tools such as gouache. The use of texture maps not only simulates a unique artistic style, but also provides a significant identifying point for patent protection, making it easy to identify technologies using similar principles and to obtain evidence.

[0031] The input image to be processed is a random value with specific distribution characteristics obtained from a texture map. This value is then used as a variable in the color change formula to achieve a non-uniform color settling effect. For example, in an image with Gaussian noise, the spatial distribution of pixel values ​​conforms to the Gaussian noise pattern. In forensic investigations, the noise texture used in the image can be reconstructed through reverse calculation. Figure 2 As shown, (1) is the rendered image, and (2) is the noise texture used to reconstruct the image by reverse calculation. By comparing the noise distribution with the standard noise distribution, it can be determined whether the image has been processed by the image rendering method of this invention.

[0032] According to the image rendering method of the present invention, in the watercolor color adjustment step, the color adjustment is performed based on the synthesized texture map, that is, only one watercolor color adjustment is needed, which reduces the amount of computation. Moreover, the synthesized texture map is a fusion of multiple texture maps, which can achieve the same rendering effect as when multiple color adjustments are performed based on multiple texture maps separately. Thus, while ensuring the rendering effect, the rendering time is reduced and the rendering efficiency is improved.

[0033] In some embodiments, the composite texture map is obtained by fusing multiple different noise texture maps; the multiple different noise texture maps are obtained by simulating different properties of watercolor paint.

[0034] Specifically, the characteristics of watercolor paints include the natural diffusion and mixing of paint in water, watermarks or paint cracking, multi-layered blending, and the graininess of the paper. Therefore, each noise texture map needs to simulate the characteristics of watercolor paints. The natural diffusion and mixing of paint in water is used to create macroscopic color changes and tones; watermarks or paint cracking are used to define structural textures and contours; multi-layered blending is used to increase the depth and complexity of the material; and the graininess of the paper is used to inject realism. After simulating the different characteristics of watercolor paints, multiple different noise texture maps are obtained. In order to efficiently render the image, the various noise texture maps are merged to obtain a composite texture map.

[0035] In some embodiments, the composite texture map is obtained by fusing Gaussian noise texture, Berlin noise texture, and paper texture.

[0036] Specifically, Gaussian noise is a type of statistical noise whose probability density function matches a Gaussian function (i.e., a normal distribution). Gaussian noise is often used to simulate random fluctuations in electronic devices or irregularities in natural phenomena. Visually, it manifests as random perturbations in the value of each pixel in an image, with the intensity following a normal distribution. Berlin noise is a gradient noise technique used in computer graphics to create a more natural and continuously varying pseudo-random pattern. Berlin noise produces a smoother, more continuous output and is well-suited for generating textures of natural phenomena such as clouds and terrain. Paper refers to a digital image obtained by scanning actual paper, including all details such as fiber structure, color variations, possible blemishes, or other imperfections.

[0037] Berlin noise textures are used to create the macroscopic structure and basic form of textures. They form the skeleton of a watercolor painting, providing a coherent, organic sense of flow. In image rendering methods, Berlin noise textures are used to simulate the blurring of watercolors, the veining of marble, or the subtle variations in the base color of an old wall.

[0038] Paper textures are used to inject realism, providing microscopic details and tactile sensations of the physical world. In image rendering methods, paper textures are used to provide the fibers of paper, subtle grayscale variations, and natural bumps and depressions.

[0039] Gaussian noise textures are used to create atmosphere, adding uniform graininess and random perturbations. As the top-level "filter," Gaussian noise textures unify the image texture and add a retro or rough feel. In image rendering methods, Gaussian noise textures are used to simulate the graininess of dust in the air or to make color transitions more natural.

[0040] For example, existing technologies use Gaussian noise, Burmester noise, and paper scanning as input textures to perform color conversion on images. This patent synthesizes three texture maps to obtain a composite texture map. For the watercolor color adjustment step, the composite texture map is used to replace the three color adjustment steps with different noise textures. Only one watercolor color adjustment is needed, which reduces the amount of computation and is suitable for vehicle-mounted deployment scenarios.

[0041] In some embodiments, image rendering further includes image abstraction processing, which includes filtering the acquired image using the Kuwabara filtering algorithm.

[0042] Specifically, image abstraction creates an artistic image with strong color contrast and simplified shapes, allowing the final result to retain the basic structure and elements of the original image while endowing it with unique artistic expression. The Kuwabara filter algorithm is a classic nonlinear image filtering algorithm, mainly used to smooth images while preserving edge information. One application of the Kuwabara filter algorithm is to achieve artistic effects reminiscent of oil paintings or watercolors; therefore, it is widely used in image abstraction.

[0043] After capturing images using a camera or webcam, the captured images are subjected to Kuwabara filtering, which preserves edges while smoothing out areas, resulting in a visual effect with distinct color blocks and low detail.

[0044] For example, in image abstraction processing, the Kuwabara filter is used instead of the existing mean-transfer filter and morphological filter. The mean-transfer filter achieves filtering through multiple iterations. Each iteration calculates the centroid of surrounding colors to update the color and offset of the current pixel, ultimately achieving color abstraction over a larger area. Morphological filtering achieves filtering by expanding the range of bright areas and reducing the range of dark areas, thereby reducing the number of intermediate colors. However, both of these filtering methods rely on multiple iterations, resulting in a large computational load and hindering optimal performance.

[0045] In some embodiments, the Kuwabara filtering algorithm includes obtaining the variance or standard deviation of pixel values ​​in each neighborhood of the target pixel of the image to be processed, wherein the iteration step size used in obtaining the variance or standard deviation is larger than the iteration step size of the standard Kuwabara filtering algorithm.

[0046] Specifically, during the Kuwabara filtering process, the target pixel in the image to be processed is determined. The target pixel can be understood as the center point of the image. The region surrounding the target pixel is divided into four quadrants. For each quadrant, the variance and standard deviation of the pixel values ​​within it are calculated, and the minimum value is selected as the new pixel result. Variance measures the degree of deviation of pixel values ​​within a region from the average value; a small variance indicates that the region's color is uniform and flat. The standard deviation, being the square root of the variance, is perhaps more intuitive. The minimum value between the variance and the standard deviation is selected as the new pixel result.

[0047] Since the color difference between adjacent pixels is not particularly large, the iteration step size used when obtaining the variance or standard deviation is larger than that of the standard Kuwabara algorithm, thus reducing the amount of computation.

[0048] For example, using the Kuwabara filtering algorithm to filter the image to be processed, the filtered image is as follows: Figure 3 As shown, by comparing the variance and standard deviation of the colors within a 6-pixel radius across the four quadrants surrounding the target pixel, the minimum value of the variance and standard deviation is selected as the new result for the pixel. In this invention, the iteration step size is changed from 1 to 2 when calculating the variance and standard deviation. For high-resolution images, this allows for a larger comparison range without increasing computational load, improving the abstraction effect and enhancing the artistic style of the image.

[0049] In some embodiments, image rendering further includes edge détenting and / or edge dithering.

[0050] Specifically, edge décor enhancement primarily aims to strengthen the contours of objects in an image, making these contours more prominent and thus increasing the image's depth and three-dimensionality. Edge dithering applies some form of perturbation to the image edges, creating an effect similar to hand-drawn or printing errors. By combining edge décor enhancement and edge dithering, the three-dimensionality of the processed image can be increased, making it more closely resemble a watercolor image.

[0051] For example, in edge darkening processing, the distance between adjacent pixels in the color space is calculated (Formula 2), which serves as a parameter in the color depth variation formula (Formula 1). The greater the distance, the darker the color becomes, thus achieving the effect of darkening the color at the edges. This invention redesigns and reimplements the formula's calculation flow using GPU algorithms, improving image processing efficiency.

[0052] C' = C(1 - (1 - C)(d - 1)); (Formula 1) Δp = |p[x-1,y] - p[x+1,y]| + |p[x,y-1] - p[x,y+1]|; (Formula 2) In Formula 1, C is the original normalized color value, d is the parameter controlling the intensity of the darkening, and C' is the darkened color value.

[0053] In Formula 2, Δp represents an approximation of the gradient magnitude at pixel (x, y); |p[x-1,y] -p[x+1,y]| is the gradient approximation in the horizontal direction (x-axis), which calculates the difference in gray values ​​between the two pixels directly to the left and right of the center pixel. If this difference is large, it indicates a drastic brightness change in the horizontal direction, i.e., there may be a vertical edge; |p[x,y-1] -p[x,y+1]| is the gradient approximation in the vertical direction (y-axis), which calculates the difference in gray values ​​between the two pixels directly above and below the center pixel. If this difference is large, it indicates a drastic brightness change in the vertical direction, i.e., there may be a horizontal edge; adding the absolute values ​​of the horizontal and vertical gradients gives the overall edge strength or degree of change Δp of the pixel. The larger the value of Δp, the higher the probability that the point is located at an edge; the smaller the value of Δp, the higher the probability that the point is located in a flat, uniformly colored area.

[0054] In some embodiments, edge jitter processing includes replacing the pixel value of each pixel in the acquired image with the pixel value at the corresponding offset position; the offset position is determined based on the random value corresponding to the pixel; the random value is assigned to each pixel of the acquired image based on the synthesized texture map.

[0055] Specifically, during edge dithering, the pixel values ​​of each pixel in the image are collected, and the pixel values ​​of each pixel in the collected image are replaced with the pixel values ​​at the corresponding offset positions. The offset positions are determined based on the random values ​​corresponding to the pixel. The synthesized texture map and the collected image overlap in the coordinate system, and their sizes can be the same or different. The noise value in the noise texture map corresponding to the pixel in the collected image can be used as the random value of that pixel. The random value is assigned to each pixel in the collected image based on the synthesized texture map. The random value is multiplied by the coordinate value of the pixel to obtain a new pixel. The new pixel can be the pixel at the offset position of the original pixel, and the pixel value of the original pixel is replaced with the pixel value at the offset position.

[0056] For example, edge jitter processing simulates brush jitter during painting by introducing a texture map to assign random values ​​to each pixel. This invention redesigns the GPU processing algorithm by using the random value of a pixel as a parameter to calculate the pixel's horizontal and vertical offset, and then replacing the current pixel value with the pixel value at the offset position, thus achieving highly efficient edge jitter simulation.

[0057] In some embodiments, the images are acquired based on images taken by a drone.

[0058] Specifically, in this invention, the images can be obtained by capturing images using a drone, such as... Figure 4As shown, images captured by the drone are transmitted to a processing module, such as a vehicle-mounted system. The processing module obtains the acquired image based on the drone's images, performs image rendering processing on the acquired image, and transforms it into unique artistic effects. It can simultaneously obtain both the acquired image and the rendered target image, resulting in clearly defined visual features. Users can readily appreciate the effectiveness of this method. Furthermore, no similar technical solutions exist from other suppliers for this image rendering method, giving it a significant technological advantage and commercial advantage in procurement and other scenarios.

[0059] In some embodiments, the image rendering method is implemented using a graphics processing unit based on the OpenGL framework.

[0060] Specifically, OpenGL (Open Graphics Library) is a cross-language, cross-platform application programming interface used to render 2D and 3D vector graphics hardware. OpenGL provides a series of functions that allow users to implement image rendering methods by writing code.

[0061] For example, the specific process of image rendering is as follows: Figure 5 As shown: Step S1, Begin.

[0062] Step S2: Load texture resources such as texture maps.

[0063] Specifically, a series of texture maps are generated using different texture map generation algorithms to simulate the different properties of watercolor paints. This step designs a texture map synthesis algorithm that can merge different texture maps into a single composite texture map, maintaining the overall texture consistency while preserving the distribution of different textures after synthesis.

[0064] Step S3: Input the footage captured by the drone.

[0065] Specifically, the encoded data stream captured by the drone is decoded and converted into RGB (Red, Green, Blue) data for input.

[0066] Step S4 involves abstracting the image to reduce its realism.

[0067] Step S5: Darken the edges to enhance color contrast at the edges.

[0068] Step S6: Add edge jitter effect to further highlight the painting features.

[0069] Step S7: Watercolor color mixing.

[0070] Specifically, different noise texture maps, such as Gaussian noise and Perrault noise, are used to simulate different distribution patterns of watercolor paint. The pixel values ​​of the noise texture maps are used as concentration, and color conversion formulas are used to calculate the variations in color intensity. Figure 6 As shown, (1) is the image before rendering, and (2) is the image after watercolor color correction.

[0071] A second aspect of this invention provides an electrical device that can implement the image rendering method of the above embodiments. In this embodiment, the electrical device may include an electronic device configured with an image rendering program. The electronic device may include, but is not limited to, a standalone device with a control unit, a standalone system-on-a-chip, or a GPU module. The electrical device may also include an electronic device with a processor and memory, a computer-readable storage medium, or a computer program product. The electrical device may include vehicles, aircraft, ferries, drones, etc., and is not specifically limited thereto.

[0072] like Figure 7 As shown, the electrical device according to an embodiment of the present invention is an electronic device including a processor and a memory. The electronic device 100 may include at least one processor 101 and a memory 102.

[0073] In this embodiment, at least one processor 101 is configured with an image rendering program; at least one processor 101 is connected to a memory 102; the memory 102 stores a computer program that can be executed by at least one processor 101, and the at least one processor 101 implements the image rendering method when executing the computer program.

[0074] According to the electronic device of the present invention, the corresponding image rendering program can be stored in the memory. When implementing the image rendering method, color adjustment is performed based on the synthesized texture map, that is, only one watercolor color adjustment is needed, which reduces the amount of computation. Moreover, the synthesized texture map is a fusion of multiple texture maps, which can achieve the same rendering effect as when multiple color adjustments are performed based on multiple texture maps separately. Thus, while ensuring the rendering effect, the rendering time is reduced and the rendering efficiency is improved.

[0075] The electrical device in this embodiment of the invention can also be a computer-readable storage medium storing a computer program, which, when executed, implements an image rendering method.

[0076] The electrical equipment in this embodiment of the invention can also be a computer program product, which includes a computer program stored on a computer-readable storage medium. The computer program includes program instructions, which, when executed by a computer, cause the computer to perform an image rendering method.

[0077] The electrical equipment in this embodiment of the invention can also be a vehicle, such as... Figure 8As shown, vehicle 200 includes electronic equipment 100.

[0078] According to the vehicle of the present invention, when rendering an image, color adjustment is performed based on a synthetic texture map, that is, only one watercolor color adjustment is required, which reduces the amount of computation. Furthermore, the synthetic texture map is a fusion of multiple texture maps, which can achieve the same rendering effect as when multiple color adjustments are performed separately based on multiple texture maps. Thus, while ensuring the rendering effect, the rendering time is reduced and the rendering efficiency is improved.

[0079] In the description of this specification, any process or method described in the flowcharts or otherwise herein may be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order according to the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0080] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0081] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0082] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0083] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0084] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

[0085] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, substrate, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example.

[0086] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. An image rendering method, characterized by, Comprising: obtaining a target image, the target image being obtained by image rendering on a to-be-processed image; the image rendering comprises a watercolor toning step, the watercolor toning step comprising toning processing based on a synthesized texture map, the synthesized texture map being obtained by fusing a plurality of texture maps.

2. The image rendering method of claim 1, wherein, the synthesized texture map is obtained by fusing a plurality of different noise texture maps; the plurality of different noise texture maps are obtained by simulating different characteristics of watercolor pigments.

3. The image rendering method of claim 1, wherein, the synthesized texture map is obtained by fusing a Gaussian noise texture, a Berlin noise texture and a paper texture.

4. The image rendering method of claim 1, wherein: the image rendering further comprises image abstraction processing, the image abstraction processing comprising filtering the captured image by a Sang-Hara filtering algorithm.

5. The image rendering method of claim 4, wherein: the Sang-Hara filtering algorithm comprises obtaining a variance or a standard deviation of pixel values in each field of a target pixel of the to-be-processed image, wherein an iteration step length used in obtaining the variance or the standard deviation is greater than an iteration step length of a standard Sang-Hara filtering algorithm.

6. The image rendering method according to any one of claims 1-5, characterized in that, the image rendering further comprises an edge deepening processing and / or an edge dithering processing.

7. The image rendering method of claim 6, wherein, the edge dithering processing comprises replacing a pixel value of each pixel point in the captured image with a pixel value of an offset position corresponding to the pixel point; the offset position is determined based on a random value corresponding to the pixel point; the random value is assigned to each pixel point in the captured image based on the synthesized texture map.

8. The image rendering method according to any one of claims 1-5, characterized in that, the captured image is obtained based on a UAV captured image.

9. The image rendering method according to any one of claims 1-5, characterized in that, the image rendering method is implemented by a graphics processing unit based on an OpenGL framework.

10. An electric device, characterized by the electrical equipment is used to implement the image rendering method of any one of claims 1-9.

Images