A method for constructing a fast HSI spatial model

Abstract

This invention belongs to the field of pattern recognition, specifically a method for constructing a fast HSI space model, enabling faster conversion between image RGB and HSI spaces. The method includes the following steps: Step 1: Obtain the image RGB space through image acquisition; Step 2: Calculate the model for converting the image RGB space to HSI space; Step 3: Calculate the model for converting the image HSI space obtained in Step 2 back to RGB space. This invention fully considers the relationship between the image RGB and HSI spaces, improving the original HSI space to achieve mutual conversion between color images in RGB and fast HSI spaces.

Description

Technical Field

[0001] This invention belongs to the field of pattern recognition, specifically a method for constructing a fast HSI space model, which enables faster conversion between image RGB space and HSI space. Background Technology

[0002] Most color images we see in daily life are encoded using the RGB color space. The RGB color space is based on three primary colors: Red (R), Green (G), and Blue (B), which are superimposed to varying degrees to produce a rich and wide range of colors; hence, it's commonly known as the three-primary-color model. However, the RGB color space is generally not used in scientific research because it represents hue, brightness, and saturation together, making them difficult to separate. Therefore, while the RGB color space is suitable for display systems, it is not suitable for image processing.

[0003] The HSI (Hue, Saturation, Intensity) color space model uses H, S, and I to describe color characteristics. H defines the frequency of a color, called hue; S represents the lightness or darkness of a color, called saturation; and I represents intensity or brightness. Hue describes the attribute of a pure color (pure yellow, orange, or red); saturation gives a measure of how much a pure color is diluted by white light; and brightness is a subjective description. The HSI color model can separate color information (hue and saturation) and illumination information (luminance component) in a color image, making it a valuable tool for developing image processing methods based on color description.

[0004] The Fast HSI Space Model is a technique for quickly converting an image from RGB space to HSI space. It uses a circular projection surface to calculate saturation and uses the cosine of the included angle to represent hue. It features fast calculation speed and low saturation. This method can significantly improve the enhancement effect of high-saturation, low-light images.

[0005] Low-light image enhancement has long been a crucial topic in pattern recognition, leading researchers to propose and develop numerous methods with significant implications and practical value. These methods can be broadly categorized into two types based on their technical characteristics: data-driven methods and model-driven methods. Data-driven methods train deep neural networks using large-scale image training, enabling them to generate normally lit images of low-light conditions. However, data-driven methods require a large number of high-quality training images, which are often difficult to obtain in practical applications. Model-driven methods, on the other hand, convert the image from RGB space to HSI space, enhance the illumination channel I, and then convert the processed HSI space image back to RGB space, achieving good enhancement results. Fast HSI space models offer high speed and low saturation. Summary of the Invention

[0006] The purpose of this invention is to provide a method for constructing a fast HSI spatial model, enabling faster conversion between RGB and HSI three channels.

[0007] The specific technical solution adopted in this invention is as follows:

[0008] A method for constructing a fast HSI spatial model includes the following steps:

[0009] Step 1: Obtain the RGB color space of the image through image acquisition. R (Red) represents the red channel of the RGB color space, G (Green) represents the green channel of the RGB color space, and B (Blue) represents the blue channel of the RGB color space. {r, g, b} represent any color point in the RGB color space, where r represents the red component value of that color point, g represents the green component value of that color point, and b represents the blue component value of that color point. The R channel of the RGB color space is composed of the r component values ​​of all color points in the image, the G channel of the RGB color space is composed of the g component values ​​of all color points in the image, and the B channel of the RGB color space is composed of the b component values ​​of all color points in the image.

[0010] Step 2: Obtain the model for converting the image from RGB space to Fast HSI space. H (Hue) represents the hue channel in the Fast HSI space, S (Saturation) represents the saturation channel, and I (Intensity) represents the lightness channel. {h, s, i} represent any color point in the Fast HSI space, where h represents the hue component value, s represents the saturation component value, and i represents the lightness component value. The H channel in the Fast HSI space is composed of the h component values ​​of all color points in the image, the S channel is composed of the s component values, and the I channel is composed of the i component values. The specific process is as follows:

[0011] Procedure 2.1: The luminance component value i of the color points in the fast HSI space of the image is defined as follows:

[0012]

[0013] Where r, g, and b are the red, green, and blue component values ​​of the color points in the RGB space of the image in step one, respectively.

[0014] Procedure 2.2: The saturation component value s of the color points in the fast HSI space of the image is defined as follows:

[0015]

[0016] Procedure 2.3: The hue component value h of the color points in the fast HSI space of the image is defined as follows:

[0017]

[0018] Equations (1), (2), and (3) are models for converting any color point {r, g, b} in the RGB space of an image into the corresponding color point {h, s, i} in the Fast HSI space of the image. Once all color points in the RGB space of the image are converted into the corresponding color points in the Fast HSI space, the H, S, and I channels of the Fast HSI space of the image can be obtained.

[0019] Step 3: Obtain the model for fast HSI space to RGB space conversion of the image obtained in Step 2. By solving equations (1), (2), and (3) in Step 2, the model for fast HSI space to RGB space conversion of the image can be obtained as follows:

[0020]

[0021]

[0022]

[0023] Equations (4), (5), and (6) convert any color point {h, s, i} in the Fast HSI space of the image into the corresponding color point {r, g, b} in the RGB space of the image. After all the color points in the Fast HSI space of the image are converted into the corresponding color points in the RGB space of the image, the R, G, and B channels of the RGB space of the image can be obtained.

[0024] The beneficial effects of this invention are as follows: This invention converts the image RGB space to HSI space. Compared with the traditional HSI space, this method has lower saturation and faster conversion speed, which is of great value in improving the performance of high-saturation low-light image enhancement. This invention addresses the separation of image color and illumination in low-light image enhancement by improving the traditional HSI space model based on a triangular projection surface to a fast HSI space model based on a circular projection surface, using the cosine of the included angle to represent hue, resulting in faster calculation speed and lower saturation. This invention can significantly improve the effect of high-saturation low-light image enhancement. Based on practical application effects, this invention offers fast calculation speed and low saturation. Attached Figure Description

[0025] Figure 1 This is a flowchart of the overall process of the method of the present invention.

[0026] Figure 2 This is a geometric relationship diagram between the RGB space and the Fast HSI space of the image. Detailed Implementation

[0027] To enhance understanding of the present invention, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. These embodiments are only used to explain the invention and do not limit the scope of protection of the invention.

[0028] Example: A method for constructing a fast HSI spatial model, the flowchart of which is as follows. Figure 1 As shown, the specific steps include:

[0029] Step 1: Obtain the RGB color space of the image through image acquisition. R (Red) represents the red channel of the RGB color space, G (Green) represents the green channel of the RGB color space, and B (Blue) represents the blue channel of the RGB color space. {r, g, b} represent any color point in the RGB color space, where r represents the red component value of that color point, g represents the green component value of that color point, and b represents the blue component value of that color point. The R channel of the RGB color space is composed of the r component values ​​of all color points in the image, the G channel of the RGB color space is composed of the g component values ​​of all color points in the image, and the B channel of the RGB color space is composed of the b component values ​​of all color points in the image.

[0030] Step 2: Obtain the model for converting the image from RGB space to Fast HSI space.

[0031] Figure 2 It is a geometric relationship diagram between the RGB space and the fast HSI space of the image. Figure 2 In the diagram, O is the origin of the three-dimensional coordinate system. Let represent the red component axis, green component axis, and blue component axis, respectively. The RGB space of an image is composed of these three color component axes. P is any color point in the RGB space of the image, and the red, green, and blue component values ​​of point P are {r, g, b}. The color point P is represented as P(r, g, b). Let O represent the vector formed by connecting points O and S. Indicates the brightness axis. The length is expressed as ΔAQF is centered at T. It is an equilateral triangle with side length . Figure 2 middle, and vertical, Perpendicular to ΔAQF, It lies in the same plane as ΔAQF.

[0032] H (Hue) represents the hue channel, S (Saturation) represents the saturation channel, and I (Intensity) represents the lightness channel in the image's Fast HSI space. {h, s, i} represent any color point in the image's Fast HSI space, where h represents the hue component value, s represents the saturation component value, and i represents the lightness component value. The H channel in the image's Fast HSI space is composed of the h component values ​​of all color points in the image; the S channel is composed of the s component values; and the I channel is composed of the i component values.

[0033] The specific process is as follows:

[0034] Step 2.1: The luminance component value i of the color points in the fast HSI space of the image is defined as follows:

[0035]

[0036] Where r, g, and b are the red, green, and blue component values ​​of the color points in the RGB space of the image in step 2.

[0037] Step 2.2: The saturation component value s of the color points in the fast HSI space of the image is defined as follows:

[0038]

[0039] in, for Figure 2 Midpoint P(r,g,b) to the brightness axis distance, yes Figure 2 The distance from the center point T of ΔAQF to vertex A.

[0040] Step 2.3: The hue component value h of the color points in the fast HSI space of the image is defined as follows:

[0041]

[0042] Where θ is Figure 2 middle and The angle between them, and 0≤θ≤180°.

[0043] Equations (1), (2), and (3) are models for converting any color point {r, g, b} in the RGB space of an image into the corresponding color point {h, s, i} in the Fast HSI space of the image. Once all color points in the RGB space of the image are converted into the corresponding color points in the Fast HSI space, the H, S, and I channels of the Fast HSI space of the image can be obtained.

[0044] Step 3: Obtain the model for converting the image from the fast HSI space to the image RGB space obtained in Step 2. By solving equations (1), (2), and (3) in Step 2, the model for converting the image from the fast HSI space to the image RGB space can be obtained as follows:

[0045]

[0046]

[0047]

[0048] Equations (4), (5), and (6) convert any color point {h, s, i} in the Fast HSI space of the image into the corresponding color point {r, g, b} in the RGB space of the image. After all the color points in the Fast HSI space of the image are converted into the corresponding color points in the RGB space of the image, the R, G, and B channels of the RGB space of the image can be obtained.

[0049] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for constructing a fast HSI spatial model, characterized in that, The method comprises the following steps: Step one, obtaining image RGB space through image acquisition; Step two, obtaining the model of image RGB space conversion to fast HSI space; Step three, obtaining the model of image fast HSI space conversion to RGB space in step two; In the step one, R represents the red channel of image RGB space, G represents the green channel of image RGB space, B represents the blue channel of image RGB space, and {r, g, b} represents any color point of image RGB space, wherein r represents the red component value of the color point, g represents the green component value of the color point, and b represents the blue component value of the color point, the R channel of image RGB space is composed of the r component value of all color points of the image, the G channel of image RGB space is composed of the g component value of all color points of the image, and the B channel of image RGB space is composed of the b component value of all color points of the image; In the step two, H represents the hue channel of image fast HSI space, S represents the saturation channel of image fast HSI space, and I represents the brightness channel of image fast HSI space, and {h, s, i} represents any color point of image fast HSI space, wherein h represents the hue component value of the color point, s represents the saturation component value of the color point, and i represents the brightness component value of the color point, the H channel of image fast HSI space is composed of the h component value of all color points of the image, the S channel of image fast HSI space is composed of the s component value of all color points of the image, and the I channel of image fast HSI space is composed of the i component value of all color points of the image; In the step two, the specific process is as follows: Process 2.1: defining the brightness component value i of the color point in the image fast HSI space; Process 2.2: defining the saturation component value s of the color point in the image fast HSI space; Process 2.3: defining the hue component value h of the color point in the image fast HSI space: After all color points of image RGB space are converted into corresponding color points of image fast HSI space, the H, S and I channels of image fast HSI space can be obtained; In the process 2.1 in the step two, the brightness component value i of the color point in the image fast HSI space is defined as follows: Wherein r, g and b are respectively the red component value, the green component value and the blue component value of the color point in the image RGB space in step one; In the process 2.2 in the step two, the saturation component value s of the color point in the image fast HSI space is defined as follows: In the process 2.3 in the step two, the hue component value h of the color point in the image fast HSI space is defined as follows: Formula (1), formula (2) and formula (3) are the model of converting any color point {r, g, b} of image RGB space into corresponding color point {h, s, i} of image fast HSI space, and after all color points of image RGB space are converted into corresponding color points of image fast HSI space, the H, S and I channels of image fast HSI space can be obtained; In the step three, by solving the formula (1), formula (2) and formula (3) in the step two, the model of the image fast HSI space conversion to the RGB space is obtained as follows: The formula (4), formula (5) and formula (6) convert the image fast HSI space any color point {h, s, i} to the image RGB space corresponding color point {r, g, b}, when all the color points of the image fast HSI space are converted to the image RGB space corresponding color points, the R, G, B three channels of the image RGB space can be obtained.

Images