Image recognition method for port district combined transportation site

Abstract

The invention relates to the technical field of image enhancement, in particular to an image recognition method for a harbor district combined transportation site, and the method comprises the steps: obtaining a target image of the harbor district combined transportation site; performing structural edge analysis on each pixel point in the target image and analysis of brightness value difference and gray value difference in a local range of the pixel point to obtain an enhanced scale control optimization factor, and adaptively obtaining an optimal multi-scale parameter combination according to the enhanced scale control optimization factor to obtain an initial enhanced image; and according to the structure maintenance optimization factor of each pixel point in the initial enhanced image, obtaining the enhancement intensity adjustment weight of each pixel point, and by using the enhancement intensity adjustment weight, carrying out compensation fusion on the brightness value of each pixel point to obtain a final brightness value so as to obtain a final enhanced image. According to the final enhanced image, an image identification task of a harbor district combined transportation site is carried out, and the image enhancement effect is improved by fully fusing double strategies of image structure perception and enhancement parameter control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image enhancement, and in particular to an image recognition method for a port area intermodal site. BACKGROUND

[0002] In port gathering and distributing intermodal operations, image acquisition and recognition technology has become a key component of intelligent management of the port area. In particular, in the night operation scene, vehicle automatic recognition, box number tracking and scene structure perception tasks have higher requirements for image quality. However, in actual application, due to the complex night lighting environment, the image acquisition of the port area often faces the problem of strong non-uniform illumination interference. Especially under the interference of high-brightness light sources such as truck headlight, large projection lamp, etc., the image is prone to problems such as saturation of strong light area, underexposure of background area, structure edge being submerged or misactivation, resulting in unbalanced image contrast and serious structure breakage, thereby seriously affecting the stability and accuracy of image enhancement and subsequent recognition tasks.

[0003] Existing image enhancement methods, especially multi-scale brightness restoration algorithms based on Retinex model, can improve the overall brightness level and local details of the image to a certain extent, but the enhancement mechanism generally lacks adaptive control of image structure content. In the strong light dominant interference area, the Retinex algorithm often preferentially enhances the high-light false edge, thereby suppressing the real structure boundary, and even causing the problem of preferential activation of edge false structure. At the same time, since Retinex enhancement mainly relies on pixel brightness gradient for filtering and reconstruction, in the structure complex area (such as characters, contours, regular object boundaries), the enhancement result is prone to destroy the structure continuity, and phenomena such as edge response breakage, texture stroke distortion, and incoherent shape occur. This enhancement imbalance problem not only damages the readability of the image, but also directly affects the recognition performance of subsequent deep learning or traditional visual algorithms.

[0004] In addition, in the scene where the image structure is severely damaged or the boundary features are intermittent, the existing enhancement technology still lacks an optimization mechanism for the whole process of "enhancement scale control, structure preservation, and recognition adaptation", and often only uses a single filter scale or uniform enhancement parameter for processing, ignoring the different needs of different regions (such as high-light disturbance area, structure edge area, and weak texture transition area) for enhancement strategies, and unable to adaptively adjust according to the actual structure distribution and response characteristics of the image, resulting in the problem that the enhancement result cannot maintain structure fidelity globally and is prone to structure cracking and false response in local, which poses a great challenge to the recognition of structural targets such as character edges and box number segments in the port area at night.

[0005] In summary, the prior art in dealing with the port night strong interference image structure enhancement task, there are mainly the following technical problems: (1) the image enhancement process lacks the identification and distinction mechanism of the structural edge and the false structure disturbance, resulting in the primary and secondary structure misjudgment problem of the enhancement result; (2) the enhancement algorithm lacks the modeling and control of the edge continuity and the structure retention ability, resulting in the problems of the target edge such as fracture and incoherent strokes. SUMMARY

[0006] Therefore, the embodiment of the present application provides an image recognition method for port intermodal site to solve the problem of poor effect of the traditional Retinex algorithm in structure enhancement of the port night strong interference image.

[0007] The embodiment of the present application provides an image recognition method for port intermodal site, which comprises the following steps:

[0008] Obtain the night operation image of the port intermodal site, preprocess the night operation image to obtain a normalized logarithmic brightness image, denoted as a target image;

[0009] Obtain the gray value and the brightness value of each pixel point in the target image, perform structural edge analysis on each pixel point in the target image and analyze the brightness value difference and the gray value difference in the local range thereof, obtain an enhancement scale control optimization factor of the target image, and adaptively obtain the optimal multi-scale parameter combination of the Retinex algorithm for image enhancement of the target image according to the enhancement scale control optimization factor, to obtain an initial enhanced image;

[0010] For any pixel point in the initial enhanced image, perform main direction structure evaluation and structure internal consistency evaluation on the any pixel point to obtain a structure retention optimization factor of the any pixel point, obtain an enhancement intensity adjustment weight of the any pixel point according to the product of the enhancement scale control optimization factor and the structure retention optimization factor, and perform compensation fusion processing on the brightness value of the any pixel point in the initial enhanced image by using the enhancement intensity adjustment weight to obtain a final brightness value of the any pixel point;

[0011] Obtain the final brightness value of each pixel point in the initial enhanced image to obtain a final enhanced image, and perform an image recognition task of the port intermodal site according to the final enhanced image.

[0012] Compared with the prior art, the embodiment of the present application has the following beneficial effects:

[0013] The application can precisely suppress non-structural false edges in areas with significant strong light interference, while realizing adaptive enhancement of real structural boundaries, effectively improving the local contrast and edge recognition of the image, and ensuring the complete preservation of structural information under complex lighting conditions. Compared with the traditional Retinex algorithm, the application can dynamically adjust the enhancement scale according to the image content, significantly reduce the over-enhancement of high-light areas and the loss of details in low-light areas, and improve the overall perception quality and foreground recognition basis of the image. Further, by constructing a structure preservation optimization factor and introducing an enhancement intensity fusion regulation mechanism, the application realizes significant optimization in image structure continuity and target recognition adaptability. The application not only can suppress edge breakage caused by brightness jump in the enhancement process, but also can dynamically adjust the enhancement intensity according to the structural direction consistency and response intensity of the pixel in the region, realizing the structure fidelity and complete depiction of key targets such as character strokes and box number edges. Overall, the application fully integrates image structure perception and enhancement parameter control dual strategies, provides a more robust perception preprocessing basis for image recognition systems in night strong interference scenes, and significantly improves the support effect of image enhancement on subsequent recognition tasks. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0015] Figure 1 is a method flowchart of an image recognition method for a port area intermodal site provided by the first embodiment of the present application. DETAILED DESCRIPTION

[0016] The embodiments of the present application will be described in detail below, and examples of the embodiments are shown in the drawings. The embodiments described below by referring to the drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application.

[0017] It should be noted that the terms "first", "second" and the like in the specification and the above drawings of the present application are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all the embodiments consistent with the present application. Rather, they are only examples of devices and methods consistent with some aspects of the present application.

[0018] In order to illustrate the technical solutions of the present application, specific examples are used for illustration below.

[0019] Referring to Figure 1 , it is a method flow chart of an image recognition method for a port area intermodal site provided by Embodiment One of the present application, as Figure 1 indicated, the method can include:

[0020] Step S101, acquiring a night operation image of the port area intermodal site, pre-processing the night operation image to obtain a normalized logarithmic luminance image, denoted as a target image.

[0021] In the port area intermodal site operation scene targeted by the present application, night image acquisition is particularly affected by complex lighting environment interference, especially when the truck enters or exits the gate or operates in the berthing area, the high-beam headlight of the truck directly irradiates the image acquisition device, which often causes strong non-uniform brightness regions in the image, showing problems such as local overexposure, blurred boundaries, and weakly exposed backgrounds. Such lighting disturbances not only destroy the overall contrast and detail distinguishability of the image, but also seriously interfere with subsequent image analysis and target recognition tasks based on structural features. In order to ensure that the subsequent enhancement and recognition tasks have a complete and stable data basis, the present application aims to adaptively plan the image acquisition process of the port area intermodal site, ensure that the input image can cover the key target area, have basic texture features, and retain enough structural information to support the subsequent development of enhancement scale control and structure preservation mechanisms.

[0022] Specifically, first, select the image acquisition position to be deployed at the key traffic nodes of the port area intermodal operation site, including truck entry and exit, gate code scanning area, waiting for loading and unloading berthing line, and dispatching guide area, etc. The image acquisition device needs to meet the high dynamic range response (HDR) capability, has night low-illumination imaging performance, and uses a fixed-view wide-angle lens to ensure that the operation targets (truck head number, box number characters, signal light state, etc.) can stably enter the field of view.

[0023] The image acquisition process is carried out in real time, with a frame rate not less than 25 fps and a resolution not less than 1080p, to ensure that the key structures (such as character strokes, truck head outline, and light boundaries) in the image have identifiable physical resolution. Each frame of acquired image contains typical high-light irradiation area, low-illumination background area and structural target area, thereby constituting a typical test sample in the subsequent image enhancement process.

[0024] After obtaining the night operation image of the port intermodal site, further basic preprocessing of the night operation image is needed to obtain the target image, and the basic preprocessing includes but is not limited to image size standardization, format conversion, gamma correction and gray scale generation and the like. It is worth noting that, in order to avoid the original brightness distribution misleading the initial state of the enhancement algorithm, the night operation image is uniformly converted into a logarithmic brightness space image, and an image normalization operation is performed before Retinex processing, so that the brightness distribution range of the input image (that is, the target image) is fixed in the [0, 1] interval, ensuring the calculation stability and comparability of the enhancement scale control optimization factor and the structure preservation optimization factor, and then obtaining the normalized logarithmic brightness image, denoted as the target image.

[0025] At this point, the target image of the port intermodal site is obtained, which is used to support the construction of the enhancement scale control optimization factor, the response of the structure preservation optimization factor and the execution of the cooperative enhancement control mechanism. At this time, not only the stable collection of the night operation image is completed, but also the initialization standardization of the target image is completed, which provides a solid foundation for the complex night port image structure enhancement task.

[0026] In step S102, the gray value and the brightness value of each pixel point in the target image are obtained, and the structural edge analysis of each pixel point in the target image and the analysis of the brightness value difference and the gray value difference in the local range thereof are performed respectively, the enhancement scale control optimization factor of the target image is obtained, and the optimal multi-scale parameter combination for performing image enhancement on the target image by using the Retinex algorithm is adaptively obtained according to the enhancement scale control optimization factor, so as to obtain an initial enhanced image.

[0027] In the night operation of the port intermodal site, vehicles frequently enter and exit, and operation scheduling is intensive. The image recognition system needs to continuously complete the automatic recognition of structural information such as license plate number and container number in a low-illumination, high-contrast and high-interference environment. However, due to objective factors such as uneven lighting conditions in the port site, frequent direct reflection of truck headlights and fixed monitoring angle, there are often local highlight areas (such as car lights and metal reflection spots) and large-area background dark parts in the collected images. In such images, the character area is often in the light edge, light transition zone or shadow area, not only the gray level is low, but also the edge structure is relatively blurred and incomplete, and the response ability of the image enhancement algorithm is limited.

[0028] The existing Retinex enhancement algorithm is usually used to adjust the brightness and enhance the details of the collected image. The algorithm estimates the illumination distribution of the image through Gaussian convolution, and calculates the reflectivity of the image in the logarithmic domain, so as to enhance the dark part and suppress the overexposure. Especially under the multi-scale Retinex framework, the algorithm processes the image at multiple scales and then fuses them to enhance the global and local details in the image at the same time. However, under the typical interference scene of night operation in the port terminal, the Retinex algorithm has the problem of misalignment of structural response capability. Since the enhancement scale parameter is fixed in the whole image, the Retinex algorithm performs the same enhancement on all regions of the image, and cannot identify the target priority of the image structure. When facing the image structure with weak character structure and strong pseudo-edge, the enhancement behavior will produce mismatched response. On the one hand, the non-structural regions such as headlight and light spot edge are significantly enlarged in the enhancement process due to their high-frequency mutation characteristics. On the other hand, the character structure cannot form an enhanced response due to its weak gray scale and fuzzy edge, and becomes more difficult to identify after enhancement.

[0029] The problem of unsuitable enhancement scale directly leads to the inconsistency between the identification region and the enhanced response region. The pseudo-edges such as light spot boundary and rust stripes often dominate the structure after enhancement, while the character region is weakened or even misjudged due to insufficient response, resulting in phenomena such as target positioning deviation, character cracking and sticking, or whole block identification failure. Therefore, the existing enhancement strategy with fixed enhancement scale cannot effectively preserve the target structure information, and will introduce significant pseudo-structure enhancement to mislead the subsequent identification task. To solve this problem, the embodiment of the present application constructs an adaptive regulation mechanism based on the structure behavior of the image at the pixel level, which is used to identify whether there is a risk of pseudo-edge dominant structure in the image, and adjusts the enhancement scale parameter accordingly. Through the joint evaluation of the structure saliency, interference response intensity and edge vulnerability at the pixel level, the real identification priority of the structure information in the current image is reflected, so that the enhancement strategy can preserve the true structure while suppressing the influence of the interference pseudo-edge.

[0030] Specifically, before performing image enhancement processing on the target image using the Retinex algorithm, the response capability of the structural target region in the image is judged to determine whether there is a risk that the structural edge is covered by the interference structure, and the enhancement scale parameter is adaptively adjusted based on this to obtain the best multi-scale parameter combination. However, for the method of obtaining the best multi-scale parameter combination, in the embodiment of the present application, each pixel point in the target image is taken as the basic calculation unit to construct the pixel-level feature response, and the enhancement scale control optimization factor required for the enhancement scale adjustment of the Retinex algorithm is constructed through coupling expression. The specific processing flow is as follows:

[0031] (1) Obtain the grayscale value and brightness value of each pixel in the target image.

[0032] For the first in the target image The nth pixel, for the nth Each pixel contributes to the grayscale response and RGB channel luminance response. The grayscale response is the grayscale value of the pixel after grayscale conversion. The RGB channel luminance response is the conversion of the three-channel pixel value of the pixel to the Y channel pixel value in the YUV color space. The conversion formula is as follows: ,in, For the first The brightness value corresponding to each pixel. The first The pixel value of each pixel in the RGB three channels.

[0033] (2) Taking any pixel in the target image as the analysis unit, structural edge analysis is performed on any pixel in the target image to obtain the structural saliency. Based on the difference in brightness value within the local range of any pixel, the degree of strong light disturbance is obtained. Based on the difference in gray value within the local range of any pixel, the degree of gray fragility is obtained.

[0034] For the first in the target image For the nth pixel, the first step is to evaluate the degree of structural edge. The horizontal and vertical grayscale gradient responses of each pixel are calculated, and the structural saliency is evaluated using the first-order derivative values ​​of the pixels. The Sobel operator is used to calculate the grayscale gradient response of each pixel in the target image. The first derivative value in the axial direction and in The first derivative value along the axis is used to obtain the sum of the absolute values ​​of the two first derivative values ​​corresponding to each pixel, which is recorded as the gray-level gradient response value of each pixel. It should be noted that the kernel size for applying the Sobel operator is set to 5×5 in this embodiment of the invention. Based on the gray-level gradient response value of each pixel in the target image, the maximum gray-level gradient response value is obtained, according to the first... The ratio between the gray-level gradient response value of the nth pixel and the maximum gray-level gradient response value yields the nth pixel. The structural significance of each pixel.

[0035] For the The formula for calculating the structural saliency of a pixel is:

[0036]

[0037] in, Indicates the first The structural saliency of each pixel; Indicates the first Pixels in The value of the first derivative in the axial direction; Indicates the first Pixels in The first derivative value in the axial direction, Represents the absolute value symbol. This represents the set of pixels in the target image. This represents extremely small positive numbers, used to prevent the denominator from being 0. In this embodiment of the invention, all extremely small positive numbers used to prevent the denominator from being 0 are specified. , This represents the maximum value function.

[0038] It should be noted that, for the first The core of evaluating the structural saliency of a pixel lies in measuring the first... Does each pixel possess sufficient edge response capability within its neighborhood? The higher the structural significance of a pixel, the more likely that the pixel is located in a structural region such as a character outline or box number line segment, and its structural information deserves to be prioritized for enhancement.

[0039] After completing the first After evaluating the structural saliency of the nth pixel, continue to evaluate the nth pixel. The strong light perturbation structure of each pixel is analyzed to identify the first pixel. Whether the region containing each pixel is located in a region dominated by pseudo-edges, such as a halo boundary or a high-contrast metallic reflection region, based on the first pixel. A local window of a preset size is constructed centered on each pixel. In this embodiment of the invention, a local window is set. The size is 5×5, used to extract the first... The local window brightness distribution of each pixel is analyzed and the local brightness contrast range is evaluated to obtain the first pixel. The intensity of strong light disturbance at each pixel: based on the first The brightness value of each pixel pair within a local window of a pixel is obtained, and the brightness value range between the maximum and minimum brightness values ​​is recorded as the target range. Similarly, the brightness value range within a local window of each pixel in the target image is obtained to get the maximum brightness value range. Based on the ratio between the target range and the maximum brightness value range, the first... The degree of strong light disturbance at each pixel.

[0040] For the The formula for calculating the intensity of strong light disturbance at each pixel is:

[0041]

[0042] in, Indicates the first The degree of strong light disturbance at each pixel; Indicates the first The maximum brightness value within a local window of a pixel. Indicates the first The minimum brightness value within a local window of a pixel. Indicates the first The maximum brightness value within a local window of a pixel. Indicates the first The minimum brightness value within a local window of a pixel. This represents the set of pixels in the target image. Represents a very small positive number. This represents the maximum value function.

[0043] It should be noted that the intensity of strong light disturbance is used to identify whether the area where the pixel is located is in a pseudo-edge dominated region such as a halo boundary or a high-contrast metallic reflection region. The higher the intensity of strong light disturbance, the more likely the pixel is in an edge region where the structure is easily disturbed by strong light. If it is not suppressed, the subsequent enhancement process will preferentially excite the pseudo-structure.

[0044] After obtaining the number After assessing the intensity of light disturbance at each pixel, we will continue to further evaluate the intensity of light disturbance at each pixel. The fragility of each pixel at the grayscale level, i.e., its contrast response capability, is determined by whether its structural edges exhibit insufficient grayscale difference due to background disturbances, uneven lighting, or other reasons, thus facing the risk of enhancement failure. The calculation then considers the vulnerability of the i-th pixel at the grayscale level. The grayscale value of the nth pixel is respectively related to the first pixel. The absolute value of the gray value difference between each pixel in a local window of n pixels is used to obtain the absolute value of the maximum gray value difference. The difference between the absolute value of the maximum gray value difference and a very small positive number is calculated. The sum of the reciprocal of the difference and a constant 1 is used as the independent variable of a logarithmic function with a base of 10 to obtain the nth pixel. The degree of grayscale fragility of each pixel.

[0045] For the The formula for calculating the grayscale fragility of a pixel is:

[0046]

[0047] in, Indicates the first The grayscale fragility of each pixel This represents a logarithmic function with base 10. Indicates the first The grayscale value of each pixel Indicates the first The grayscale value of each pixel Represents a very small positive number. Indicates the first A local window of 1 pixel.

[0048] Similarly, according to the above... The analysis method for each pixel is used to obtain the structural saliency, strong light disturbance degree, and grayscale fragility of each pixel in the target image.

[0049] (3) Obtain the structural saliency, strong light disturbance degree and gray level fragility of each pixel in the target image, and combine the structural saliency, strong light disturbance degree, gray level fragility, gray level value and brightness value of each pixel in the target image to obtain the enhancement scale control optimization factor of the target image.

[0050] Specifically, after obtaining the structural saliency, strong light perturbation level, and grayscale fragility of each pixel, the three pixel-level analyses are fused and evaluated to obtain the enhancement scale control optimization factor. The calculation expression for the enhancement scale control optimization factor is as follows:

[0051]

[0052] in, This represents the enhancement scale control optimization factor of the target image. Indicates the first in the target image The structural saliency of each pixel Indicates the first in the target image The degree of strong light disturbance at each pixel Indicates the first in the target image The grayscale fragility of each pixel This indicates that the first element in the target image obtained using the Sobel operator is... Brightness gradient of each pixel This indicates that the Laplacian operator is used to calculate the first value in the target image. The second derivative of each pixel is calculated, and the kernel size of the Laplacian operator is also set to 5×5. The set of pixels in the target image is represented by ||, where | represents the absolute value symbol and 1 represents a constant.

[0053] It should be noted that the numerator in the formula for calculating the enhanced scale control optimization factor is composed of the product of structural salience, the degree of strong light perturbation, and the degree of gray-level fragility. Among these factors, structural salience... Used to measure the The edge strength of each pixel, through and First-order gray-level gradient evaluation in two directions Whether a pixel is located in a structural region with a clearly defined edge shape, such as an outline, line segment, or character edge, is amplified by squaring the weight of structural edge regions, making them dominant in the overall enhancement dynamics; the degree of strong light disturbance... Used to evaluate the Whether the local area containing each pixel is located in a region with extremely uneven brightness, such as a halo boundary, a high-contrast metallic reflective band, or other pseudo-edge dominated areas; the higher the degree of strong light disturbance, the more it indicates that the pixel is located in a region with extremely uneven brightness, such as a halo boundary, a high-contrast metallic reflective band, or other pseudo-edge dominated areas. The more likely an area is to be dominated by lighting interference; grayscale fragility. Evaluate the first from the perspective of contrast If a pixel is located in a weak edge region with indistinct structure, and the local grayscale change is not significant (i.e., the grayscale difference is small), it indicates that the structure of the region is weak and more prone to distortion. During enhancement, additional contrast stretching is required. Therefore, the product of these three factors constitutes a risk signal for image enhancement. If a pixel simultaneously possesses structural edge features, is located in an area subject to strong light interference, and has its own grayscale response risk, then the need for enhancement scale for this pixel is particularly urgent, and a significant enhancement response should be given.

[0054] In the denominator of the formula for calculating the optimization factor for enhanced scale control, This represents the inverse ratio factor for height gradient stability, used to evaluate the degree of spatial variation in the brightness dimension of an image. If the brightness gradient of a pixel is large, i.e., the brightness is unstable, then... If the value is small, and the brightness area of ​​the pixel is stable, then The value is relatively large. This term acts as an amplifier in the denominator to punish excessive enhancement requirements. When the overall brightness of the image changes drastically, it indicates that the area may be under complex lighting conditions, and the enhancement scale should not be increased blindly. The image structure complexity factor, obtained by the second derivative of the Laplacian operator, reflects the intensity of changes in the local structure and the complexity of the texture. The more complex the structure, the more detailed the image itself is, and it should not be over-enhanced to avoid destroying the original information. Therefore, this term also serves as a stability constraint to suppress the enhancement scale. The denominator in the calculation formula of the enhancement scale control optimization factor reflects the dual suppression effect of structure and brightness complexity, ensuring that the enhancement process will not cause an overreaction in areas that already have cleaned structures.

[0055] The enhancement scale control optimization factor is used to construct a whole image scale control factor according to the analysis of each pixel point in a manner of driving the whole by the local and feeding back the global, so as to enhance and stimulate the area with dominant interference, weak structure and insufficient contrast, and to enhance and inhibit the area with saturated structure and complex texture. Therefore, after obtaining the enhancement scale control optimization factor of the target image, the multi-scale filter set in the Retinex processing process is optimized by the enhancement scale control optimization factor, that is, the optimal multi-scale parameter combination for image enhancement of the target image by the Retinex algorithm is adaptively obtained according to the enhancement scale control optimization factor, and an initial enhanced image is obtained: a preset multi-scale parameter combination for a structure clear image and a preset multi-scale parameter combination for an interference dominant image are obtained, the enhancement scale control optimization factor is taken as the weight of the preset multi-scale parameter combination for the interference dominant image, a difference between a constant 1 and the enhancement scale control optimization factor is taken as the weight of the preset multi-scale parameter combination for the structure clear image, the preset multi-scale parameter combination for the structure clear image and the preset multi-scale parameter combination for the interference dominant image are weighted and summed, and an optimal multi-scale parameter combination is obtained; the target image is image-enhanced by the Retinex algorithm based on the optimal multi-scale parameter combination, and an initial enhanced image is obtained.

[0056] In an embodiment, a multi-scale parameter combination suitable for a structure clear image is set , a multi-scale parameter combination suitable for an interference dominant image is set , an adaptive optimal multi-scale parameter combination is obtained in a manner of , and the target image is enhanced by the Retinex algorithm through the adaptive optimal multi-scale parameter combination to obtain an initial enhanced image, wherein the Retinex algorithm belongs to the prior art and is not described in detail here.

[0057] Up to now, the target image is enhanced for the first time through the joint analysis of the structure edge and the strong light interference response of the image pixel level, and an initial enhanced image is obtained.

[0058] In step S103, for any pixel point in the initial enhanced image, the structure direction of any pixel point is evaluated and the structure internal consistency of any pixel point is evaluated, a structure retention optimization factor of any pixel point is obtained, an enhancement intensity adjustment weight of any pixel point is obtained according to the product of the enhancement scale control optimization factor and the structure retention optimization factor, the brightness value of any pixel point in the initial enhanced image is compensated and fused by using the enhancement intensity adjustment weight, and a final brightness value of any pixel point is obtained.

[0059] In the process of night operation in the port terminal, the image acquisition is often disturbed by strong non-uniform lighting, especially when the truck headlight directly shoots the acquisition device, which will cause a significant brightness saturation area in the image, while the background area is still in a low-illumination state, and the overall image presents a strong dynamic uneven light distribution phenomenon. In order to improve the availability of such images in subsequent target identification tasks, the foregoing step S102 has suppressed the false edges caused by strong light interference by constructing an enhanced scale control optimization factor based on edge structure consistency, and enhanced the structural true boundary, thereby improving the overall image structure definition and identification foreground quality. However, after completing the enhanced scale control optimization, further analysis of the initial enhanced image can find that, although the enhanced scale has realized dynamic adjustment, due to the Retinex algorithm itself does not have a structure keeping mechanism in the process of multi-scale weighting, there are still the following problems: first, part of the image in the true structure internal region (such as the internal character stroke, target texture region) is over-enhanced in the enhancement process, which makes the originally uniform structure region appear intensity disturbance, thereby introducing non-real boundaries in the subsequent identification; second, there are enhancement response interruptions at some edges due to brightness jump, which makes the true structure edge present fragmentation or fragmentation, and then affects the continuity of character contour identification or target shape modeling.

[0060] The essence of the above problem is not from the enhanced scale control itself, but from the lack of structure continuity evaluation ability of Retinex enhancement, especially in the structure transition area under extreme lighting, which is easy to cause non-structure consistent response in the enhancement result due to the local feature being too sensitive. Therefore, in order to ensure that the structure integrity of the image can still be effectively repaired and maintained after the enhanced scale control optimization, the embodiment of the present application introduces a set of optimization mechanism specially for structure response keeping, so that the enhancement process not only has region selection, but also needs to have structure continuity control ability.

[0061] From an algorithmic perspective, the Retinex algorithm, as an image enhancement method based on a reflectivity and illumination separation model, relies heavily on local brightness gradients and multi-scale weighting strategies in edge regions for its enhancement effect. However, it lacks modeling of macroscopic structural information such as structural continuity, edge closure, and stroke preservation. Under extreme lighting conditions, edge continuity and internal structural consistency often cannot be reflected simply by pixel local responses. Therefore, it is necessary to introduce modeling and adjustment of structural response patterns in addition to the enhancement mechanism. Further analysis reveals that the areas most severely affected by strong light disturbance in nighttime operation images are often not the target edges themselves, but rather the surrounding areas with weak structural responses and discontinuous transition zones. These areas are prone to response drift or abrupt changes under Retinex enhancement, becoming the cause of structural breakage. In summary, based on the already completed enhancement scale control optimization, the embodiments of the present invention will further optimize the structural response continuity control. By combining the response characteristics of the real structure in the image and the heterogeneous behavior of the edge transition zone, the continuous edge expression ability of each pixel in the main direction of the structure is evaluated, and combined with its response consistency in the internal region of the structure, a response regulation mechanism for fine adjustment of structural closure and stroke integrity is finally formed. This further improves the structural fidelity and the overall stability of image recognition without interfering with the previous enhancement effect.

[0062] For the initial augmented image, the dominant orientation structure of each pixel is evaluated. The Sobel operator is used to perform gradient direction analysis on the image to obtain the dominant structural orientation of each pixel, taking the first pixel in the initial augmented image as an example. Taking the nth pixel as an example, for the nth pixel... The dominant direction of the structure of each pixel ,in, Indicates the first Pixels in Gradient value in the axial direction, Indicates the first Pixels in Gradient value in the axial direction, This represents the arctangent function.

[0063] Based on the The dominant structural direction of each pixel is used to define the sampling band length. In this embodiment of the invention, the sampling band length is set. In the A one-dimensional edge response sampling band aligned with the main direction is constructed around each pixel. This sampling band is used to evaluate structural continuity in the main direction. Therefore, based on the gradient value of each pixel in the sampling band, the absolute value of the gradient value difference between every two adjacent pixels is calculated, and the mean of the absolute values ​​of the gradient value differences is obtained. The gradient value of each pixel in the sampling band is then compared with the gradient value of the first pixel. The absolute value of the difference between the gradient values ​​of the nth pixel is used to obtain the maximum absolute value of the difference. This maximum absolute value is used as the numerator, and the gradient values ​​of the nth pixel are used as the numerator. The ratio is obtained by using the sum of the gradient values ​​of the nth pixel and the smallest positive number as the denominator. The sum of the average absolute values ​​of the gradient value differences and the ratios is then used to obtain the nth pixel. The degree of discontinuity of each pixel.

[0064] For the The formula for calculating the degree of discontinuity of a pixel is:

[0065]

[0066] in, Indicates the first The degree of discontinuity of each pixel Indicates the length of the sampling band. Indicates the first The sampling band of the nth pixel Gradient values ​​of each pixel Represents the maximum value function. Indicates the first The sampling band of the nth pixel Gradient values ​​of each pixel Represents a very small positive number. Indicates the first Gradient values ​​of each pixel || represents the absolute value of the maximum difference, and || represents the absolute value sign.

[0067] It should be noted that, The degree of discontinuity is used to measure the first The fluctuation and abrupt change of the edge response of each pixel in its main direction sampling band are used to measure whether there are breaks, jumps, or discontinuities in the edge. Let be the average difference fluctuation term, representing the th The magnitude of gradient response variation among adjacent pixels in the main direction sampling band of the nth pixel indicates the magnitude of the average difference fluctuation term. The discontinuous fluctuations in the main direction of each pixel may be due to unnatural phenomena such as structural breaks or edge splits. The maximum mutation penalty term represents the term of the th mutation. The maximum difference in gradient between the nth pixel and all points in the main direction sampling band is used to explicitly penalize abrupt changes in structural edges. If there is a significant jump between a pixel and its surrounding pixels, it indicates that the pixel is at a structural discontinuity and should be assigned a higher degree of discontinuity. The greater the degree of discontinuity in the overall structure, the higher the degree of discontinuity. The edges of individual pixels are discontinuous and abrupt in their main direction, making them suitable for detecting problem areas such as character edge interruption, stroke breakage, and missing box number outline.

[0068] After obtaining the number After determining the discontinuity of the first pixel, continue with the second pixel. The structural internal consistency response of the nth pixel is evaluated to determine the nth pixel. Whether a pixel is located within a structurally closed region (such as the center of a character stroke or the interior of a shape) is determined by using the first pixel as an example. The structural response mean and orientation consistency within a local window centered on pixel n are used to determine the first... The evaluation of the structural internal consistency response of each pixel is specifically based on the following: A local window of a preset size is constructed centered on each pixel. In this embodiment of the invention, a local window is set. The size is 5×5. Any pixel in the local window is taken as the target pixel. The dominant structural direction of the target pixel and the sampling band along that direction are obtained. A directional high-pass kernel is used to filter the sampling band of the target pixel to obtain the filtered grayscale value of the target pixel. This value characterizes whether there is a structural enhancement response in the image along its dominant structural direction at the target pixel, i.e., whether the image has a local consistency response of structural targets such as strokes, contours, and closed blocks. The relationship between the target pixel and the first... The structural dominance direction similarity between the nth pixel is obtained based on the filtered grayscale value and the structural dominance direction similarity, thus showing the relationship between the target pixel and the nth pixel. The structural orientation consistency index between pixels is obtained; the relationship between each pixel within the local window and the first pixel is also obtained. The structural orientation consistency index among all pixels is used as the mean of the structural orientation consistency indexes. The degree of structural internal consistency response of each pixel.

[0069] For the The formula for calculating the internal structural consistency response of a pixel is:

[0070]

[0071] in, Indicates the first The degree of structural internal consistency response of each pixel; Indicates the first A local window of 1 pixel, Indicates the first In the local window of the nth pixel, the first The filtered grayscale value of each pixel; Indicates the first In the local window of the nth pixel, the first The structure of each pixel dominates the direction. Indicates the first The structure of each pixel dominates the direction. Indicates the first The number of pixels within a local window of a given pixel. This represents the cosine function.

[0072] It should be noted that, The degree of internal consistency of the structure is used to measure the first To determine whether a pixel is located within a structurally defined internal region (such as the center of a character or inside a texture block), it is necessary to ascertain whether the structural orientation of that region is consistent and evaluate whether it constitutes a true structurally closed region. Indicates the first In the local window of the nth pixel, the first Does each pixel exhibit structural enhancement responses (e.g., contours, closed regions) along its principal structural direction? Convolution with a structural direction band filter yields a detection function capable of detecting stroke responses. For directional consistency assessment, measure the first Whether the structural orientation of a pixel is consistent with that of its neighboring pixels. If the directional difference is small, the value is close to 1, indicating that there is a consistent structural response within the region, which may belong to a closed stroke, region filling, or inside a boundary. The higher the internal structural consistency response of the nth pixel, the more it indicates that the nth pixel is in good condition. The more a pixel is located in a region with consistent structural orientation and strong response, the higher the probability that it is inside the structure. It should not be over-enhanced or misjudged as an edge, thus providing a basis for judgment on structural protection and response maintenance.

[0073] After obtaining the number After assessing the discontinuity of the first pixel and the internal consistency of the structure, we continue by... The discontinuity of each pixel and the internal consistency response of the structure are fused and evaluated to obtain the structure preservation optimization factor of the pixel: the product of the discontinuity of any pixel and the internal consistency response of the structure is obtained, and the reciprocal of the sum of the constant 1 and the product is used as the structure preservation optimization factor of any pixel.

[0074] For the The formula for calculating the structure preservation optimization factor for each pixel is:

[0075]

[0076] in, Indicates the first The structure of each pixel is maintained by an optimization factor; Indicates the first The degree of discontinuity of each pixel; Indicates the first The degree of structural internal consistency response of each pixel, where 1 represents a constant.

[0077] It should be noted that the optimization factor is maintained through structure. Conduct a fusion assessment, if the first If both the discontinuity and structure preservation optimization factor of a pixel are high, it indicates that the enhancement caused a structural interruption, and the break point is in a critical structural region; the enhancement response intensity should be significantly reduced. If the discontinuity is high but the structure preservation optimization factor is low, it indicates that although there is an interruption, it is not within the structural region and may be a weak background structure; the restrictions can be appropriately relaxed. If the discontinuity is low but the structure preservation optimization factor is high, it indicates that the internal structure has strong continuity but no abrupt changes; the current response can be maintained. If both are low, the structural risk in this region is considered low, and no intervention is needed. The value of the structure preservation optimization factor is... The larger the value, the higher the degree of structural preservation, and the more relaxed the adjustment intensity can be. Conversely, the smaller the value, the higher the structural protection requirement, and the stronger the constraints should be.

[0078] Similarly, a structure preservation optimization factor is obtained for each pixel in the initial enhanced image. This factor is used to correct the structural response breaks and stroke edge blurring caused by the failure to evaluate structural continuity during Retinex enhancement, thereby improving the coherence and integrity of image structural elements in subsequent recognition tasks. At this point, a joint analysis of the character edge direction coherence and local structural response intensity in the initial enhanced image is completed. To further improve the structural continuity and edge closure of the enhanced image, a strong scale control optimization factor for the target image is established. And maintain the optimization factor of the structure of each pixel. The enhancement results are fused and a fusion enhancement control weight is introduced in the Retinex enhancement output stage to perform pixel-level compensation adjustment, thereby achieving dynamic control of enhancement intensity under structure perception.

[0079] Specifically, for the first in the initial enhanced image Each pixel is used to construct an enhanced intensity adjustment weight. To balance the degree of fusion between the Retinex enhancement result and the original image brightness, the enhancement scale control optimization factor and the first... The structure of each pixel maintains the product of the optimization factors, which serves as the product of the optimization factors for the nth pixel. The enhancement intensity adjustment weights for each pixel are as follows. The formula for calculating the enhancement intensity adjustment weight of each pixel is:

[0080]

[0081] in, Indicates the first The enhancement intensity of each pixel is adjusted by weight; The enhancement scale control optimization factor represents the target image; Indicates the first in the initial enhanced image The structure of each pixel is maintained by an optimization factor.

[0082] Based on the enhancement intensity adjustment weights, the Retinex enhancement results are subjected to structure-preserving compensation fusion processing to obtain the first-order enhancement image in the initial enhanced image. The final brightness value of the nth pixel: Get the nth pixel's brightness value. The brightness value of the nth pixel in the initial enhanced image is denoted as the enhanced brightness value. This value is then compared to the brightness value of the nth pixel in the target image. The brightness values ​​of pixels belonging to the same position are recorded as the original brightness values. The enhancement intensity adjustment weight is used as the weight for the enhanced brightness value, and the difference between the constant 1 and the enhancement intensity adjustment weight is used as the weight for the original brightness value. The enhanced brightness value and the original brightness value are then weighted and summed to obtain the first... The final brightness value of each pixel.

[0083] For the The formula for calculating the final brightness value of each pixel is:

[0084]

[0085] in, Indicates the first The final brightness value of each pixel. Indicates the first Adjusting the enhancement intensity weight for each pixel Indicates the first in the initial enhanced image The brightness value of each pixel. Indicates the target image with the first The brightness values ​​of pixels at the same position.

[0086] It should be noted that the optimization factor is controlled through global enhanced scaling. Unify the constraint enhancement range and combine local structure-preserving optimization factors. A dynamic weighted evaluation of the structural coherence of specific pixels is performed, ensuring that the enhanced output retains strong responsive structures while suppressing unstructured enhancements caused by discontinuous edges or internal texture perturbations. This is achieved by introducing... The Retinex enhancement result is no longer a direct enhancement output in a single scale, but an enhancement result with controllable structure after optimization via structure perception. For pixel points in a high-structure reliable area such as a character contour or a box number edge, the value is usually high, the value formed after superposition is also large, and the enhancement result can be fully retained; and for pixel points disturbed by strong light or having an excessively large internal disturbance response, the value is small, and the final enhancement output tends to be conservative fusion, thereby significantly improving the recoverability of the structure integrity and the overall recognition stability of the enhanced image.

[0087] Similarly, the final luminance value of each pixel point in the initial enhanced image is obtained.

[0088] In step S104, the final luminance value of each pixel point in the initial enhanced image is obtained, and a final enhanced image is obtained. The image recognition task of the port intermodal site is performed according to the final enhanced image.

[0089] After obtaining the final luminance value of each pixel point in the initial enhanced image, the final luminance value is used to replace the luminance value of each pixel point in the initial enhanced image, and a final enhanced image is obtained, thereby completing the structure enhancement and light self-adaptive regulation of the image. Then, according to the actual needs of the typical recognition objects of the port intermodal site, the final enhanced image is further subjected to structured feature extraction and image recognition processing, so as to realize high-precision positioning and information extraction of the key recognition target. Since the final enhanced image has realized suppression of the pseudo-edge dominant area and enhancement of the real structure boundary, the final enhanced image has clear character structure, good stroke coherence and uniform local contrast, thereby laying a high-quality input foundation for the subsequent recognition process. In the port intermodal operation scene, typical image recognition objects mainly include container numbers, truck license plates, company logos, personnel postures, intermodal signal light states and the like. These targets are generally located in high-reflective areas of the vehicle body, low-illumination backgrounds at night or complex shielding environments, and have characteristics such as large size variation, blurred target edges and many structural disturbances. To cope with the above challenges, the embodiment of the present application first extracts high-contrast regions with character morphology in the final enhanced image by means of MSER region stability detection or EAST text detection network, and forms a character candidate region set. Under the Retinex structure enhancement processing, the quality of the character candidate regions in the character candidate region set is obviously improved, and the false detection rate is greatly reduced.

[0090] ​​​​Subsequently, in combination with image geometric structure analysis, the character candidate regions of the character candidate region set are analyzed in terms of structural indicators such as aspect ratio, edge density, stroke thickness, etc., to complete character line aggregation and redundant region elimination. Through this process, the original character candidate regions are regularized into character regions with recognition significance. Further, to enhance recognition robustness, the extracted character regions are subjected to affine correction, grayscale normalization and stroke smoothing,

[0091] After completing the structured feature extraction and normalization processing, the recognition model stage adopts existing high-performance target recognition and character recognition networks for end-to-end recognition processing. For the character recognition task, Tesseract OCR or a deep learning character recognition model based on the CRNN structure can be selected to recognize printed characters such as container numbers, license plate numbers and company logos; for the multi-target detection task such as intermodal signals and personnel actions, YOLOv7 or PP-YOLO series networks can be selected for real-time target detection; for the personnel behavior state recognition task, human key point detection methods such as OpenPose or AlphaPose can be combined to recognize the gestures or standing positions of operators, and accordingly infer the operation instructions or operation process status.

[0092] Finally, the recognition results are structured and output to generate a recognition result set including vehicle identity, container number, intermodal status, personnel posture and other information, and are simultaneously pushed to the intermodal dispatching system or security control platform for subsequent process linkage or abnormal event warning. It should be noted that the focus of the present application is on image enhancement of night operation images, and the image recognition task in the intermodal site of the port area belongs to the prior art, which will not be described in detail here.

[0093] The above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for image recognition for a port container yard, characterized in that, The method comprises: obtaining a night operation image of a port intermodal terminal site, preprocessing the night operation image to obtain a normalized logarithmic luminance image, denoted as a target image; obtaining the gray value and the luminance value of each pixel point in the target image, performing structural edge analysis on each pixel point in the target image and analyzing the luminance value difference and the gray value difference in the local range thereof, obtaining an enhanced scale control optimization factor of the target image, and adaptively obtaining an optimal multi-scale parameter combination for image enhancement of the target image by using a Retinex algorithm according to the enhanced scale control optimization factor to obtain an initial enhanced image; for any pixel point in the initial enhanced image, performing main direction structure evaluation and structure internal consistency evaluation on the any pixel point to obtain a structure preservation optimization factor of the any pixel point, obtaining an enhanced intensity adjustment weight of the any pixel point according to the product of the enhanced scale control optimization factor and the structure preservation optimization factor, and performing compensation fusion processing on the luminance value of the any pixel point in the initial enhanced image by using the enhanced intensity adjustment weight to obtain a final luminance value of the any pixel point; obtaining the final luminance value of each pixel point in the initial enhanced image to obtain a final enhanced image, and performing an image recognition task of the port intermodal terminal site according to the final enhanced image.

2. The image recognition method for a port-related intermodal site according to claim 1, wherein, The method comprises: for any pixel point in the target image, performing structural edge analysis on the any pixel point to obtain a structure saliency degree, obtaining a strong light disturbance degree according to the luminance value difference in the local range of the any pixel point, and obtaining a gray vulnerability degree according to the gray value difference in the local range of the any pixel point; obtaining the structure saliency degree, the strong light disturbance degree and the gray vulnerability degree of each pixel point in the target image, and combining the structure saliency degree, the strong light disturbance degree, the gray vulnerability degree, the gray value and the luminance value of each pixel point in the target image to obtain the enhanced scale control optimization factor of the target image.

3. The image recognition method for a port-related intermodal site according to claim 2, wherein The method comprises: The first derivative value of each pixel point in the target image in the x direction and the first derivative value of each pixel point in the y direction are calculated respectively by using a Sobel operator. The sum of the absolute values of the two first derivative values corresponding to each pixel point is obtained, and is recorded as the gray gradient response value of each pixel point.​ obtaining the structure saliency degree of the any pixel point according to the ratio between the gray gradient response value of the any pixel point and the maximum gray gradient response value.

4. The image recognition method for a port-related intermodal site according to claim 2, wherein The method comprises: obtaining the strong light disturbance degree according to the luminance value difference in the local range of the any pixel point. A local window of a preset size is constructed with the any pixel point as a center, a brightness value range between a maximum brightness value and a minimum brightness value is obtained according to brightness values of each pixel point pair in the local window, and the brightness value range is recorded as a target range, a brightness value range in a local window of each pixel point in the target image is obtained, and a maximum brightness value range is obtained, and a strong light disturbance degree of the any pixel point is obtained according to a ratio between the target range and the maximum brightness value range.

5. The image recognition method for a port-related transshipment site according to claim 4, wherein The gray scale vulnerability degree is obtained according to a gray scale value difference in a local range of the any pixel point, and the gray scale vulnerability degree comprises: A maximum gray scale value difference absolute value is obtained by calculating gray scale value difference absolute values between the gray scale value of the any pixel point and gray scale values of each pixel point in the local window of the any pixel point, and a difference value between the maximum gray scale value difference absolute value and a minimum positive number is calculated, and a sum of an inverse of the difference value and a constant 1 is taken as an independent variable of a logarithmic function with a constant 10 as a base, and a gray scale vulnerability degree of the any pixel point is obtained.

6. The image recognition method for a port-related intermodal site according to claim 2, wherein The enhanced scale control optimization factor of the target image is obtained by combining the structure saliency degree, the strong light disturbance degree, the gray scale vulnerability degree, the gray scale value and the brightness value of each pixel point in the target image, and the enhanced scale control optimization factor comprises: ； in, This represents the enhancement scale control optimization factor of the target image. Indicates the first in the target image The structural saliency of each pixel Indicates the first in the target image The degree of strong light disturbance at each pixel Indicates the first in the target image The grayscale fragility of each pixel This indicates that the first element in the target image obtained using the Sobel operator is... Brightness gradient of each pixel This indicates that the Laplacian operator is used to calculate the first value in the target image. The second derivative of each pixel The set of pixels in the target image is represented by ||, where | represents the absolute value symbol and 1 represents a constant.

7. The image recognition method for a port-related intermodal site according to claim 1, wherein The optimal multi-scale parameter combination for image enhancement of the target image by using the Retinex algorithm is adaptively obtained according to the enhanced scale control optimization factor, and an initial enhanced image is obtained, and the optimal multi-scale parameter combination comprises: A preset multi-scale parameter combination for a structure clear image and a preset multi-scale parameter combination for an interference dominant image are obtained, the enhanced scale control optimization factor is taken as a weight of the preset multi-scale parameter combination for the interference dominant image, a difference value between a constant 1 and the enhanced scale control optimization factor is taken as a weight of the preset multi-scale parameter combination for the structure clear image, and a weighted sum of the preset multi-scale parameter combination for the structure clear image and the preset multi-scale parameter combination for the interference dominant image is obtained, and the optimal multi-scale parameter combination is obtained. The target image is image-enhanced by using the Retinex algorithm based on the optimal multi-scale parameter combination, and the initial enhanced image is obtained.

8. The image recognition method for a port-related intermodal site according to claim 1, wherein, The structure retention optimization factor of the any pixel point is obtained by performing main direction structure evaluation and structure internal consistency evaluation on the any pixel point, and the structure retention optimization factor comprises: A structure dominant direction of the any pixel point is obtained, a sampling band of a preset length of the any pixel point is obtained in the structure dominant direction, a gradient value difference absolute value between each two adjacent pixel points is calculated according to gradient values of each pixel point in the sampling band, a mean value of the gradient value difference absolute values is obtained, a difference absolute value between a gradient value of each pixel point in the sampling band and a gradient value of the any pixel point is calculated, a maximum difference absolute value is obtained, a corresponding ratio is obtained by taking the maximum difference absolute value as a numerator and a sum of the gradient value of the any pixel point and a minimum positive number as a denominator, and an interruption degree of the any pixel point is obtained according to an addition value between the mean value of the gradient value difference absolute values and the ratio. A local window of a preset size is constructed with the any pixel point as a center, any pixel point in the local window is taken as a target pixel point, a structure dominant direction of the target pixel point and a sampling band in the structure dominant direction are obtained, a filtering operation is performed on the sampling band of the target pixel point, a filtered gray value of the target pixel point is obtained, a structure dominant direction similarity between the target pixel point and the any pixel point is calculated, a structure direction consistency index between the target pixel point and the any pixel point is obtained according to the filtered gray value and the structure dominant direction similarity; a structure direction consistency index between each pixel point in the local window and the any pixel point is obtained, and an average of all structure direction consistency indexes is taken as a structure internal consistency response degree of the any pixel point; The structure keeping optimization factor of the any pixel point is obtained by combining the discontinuity degree and the structure internal consistency response degree of the any pixel point.

9. The image recognition method for a port-related transshipment site according to claim 8, wherein, The structure keeping optimization factor of the any pixel point is obtained by combining the discontinuity degree and the structure internal consistency response degree of the any pixel point, including: A product of the discontinuity degree and the structure internal consistency response degree of the any pixel point is obtained, and an inverse of an addition value of a constant 1 and the product is taken as the structure keeping optimization factor of the any pixel point.

10. The image recognition method for a port-related intermodal site according to claim 1, wherein, The final luminance value of the any pixel point is obtained by performing compensation fusion processing on the luminance value of the any pixel point in the initial enhanced image by using the enhanced intensity adjustment weight, including: The luminance value of the any pixel point in the initial enhanced image is obtained and is denoted as an enhanced luminance value, the luminance value of a pixel point belonging to the same position as the any pixel point in the target image is obtained and is denoted as an original luminance value; the enhanced intensity adjustment weight is taken as a weight of the enhanced luminance value, a difference value of a constant 1 and the enhanced intensity adjustment weight is taken as a weight of the original luminance value, and the enhanced luminance value and the original luminance value are weighted and summed to obtain the final luminance value of the any pixel point.

Images