Low-altitude intelligent integrated rail transit system based on BIM technology

Abstract

The present invention relates to the field of image enhancement technology, and specifically to a low-altitude intelligent fusion rail crossing integrated system based on BIM technology. It includes an acquisition module, which is used to acquire the inspection grayscale image and divide it into different equal blocks; the area division obtains the normal area and abnormal area affected by the lighting; the detail recognition module is used to perform edge detection to obtain the edge pixel points in each area, and determine the influence degree of the detail features of each abnormal area; the distribution analysis module is used to determine the distribution concentration degree of the abnormal area according to the position distribution characteristics of all abnormal areas; the enhancement module is used to determine the influence degree of lighting according to the influence degree and distribution concentration degree of the detail features of each abnormal area; according to the influence degree of lighting, the lighting influence area to be enhanced is screened out, and the image enhancement is performed on the lighting influence area to be enhanced. The present invention can enhance the clarity of image details and improve the quality of the enhanced image.

Description

Technical Field

[0001] The present invention relates to the field of image enhancement technology, and in particular to a low-altitude intelligent fusion rail crossing integration system based on BIM technology. Background Art

[0002] The combination of drones and BIM technology has significantly improved the efficiency and accuracy of low-altitude intelligent integrated rail transit inspections. In related technologies, by configuring image acquisition equipment on drones, large areas can be quickly covered, and high-definition images and data can be obtained in real time, reducing the time and labor costs required for manual inspections. By accurately detecting minor defects in infrastructure and connecting data with BIM models, potential problems can be discovered in a timely manner.

[0003] The image acquisition device has a certain image adaptive adjustment function, such as automatically reducing the sensitivity in areas with strong light and automatically increasing the amount of light received in areas with darker light. However, this type of adjustment is limited to the overall enhancement of the video based on the current scene. When low-altitude inspections encounter complex scenes, for example, most areas of the overall image are in a normal light-sensitive state, while some areas are affected by strong light or backlight. At this time, unified image enhancement will result in poor optimization effect, which will make the image details unclear and the image quality low. Summary of the invention

[0004] In order to solve the technical problem that the optimization effect of image uniform enhancement in the related technology is too poor, resulting in unclear image details and low image quality, the present invention provides a low-altitude intelligent fusion rail transit integration system based on BIM technology, and the technical solution adopted is as follows:

[0005] The present invention proposes a low-altitude intelligent fusion rail transit integration system based on BIM technology, including:

[0006] An acquisition module is used to acquire inspection grayscale images of different frames during the low-altitude inspection process, and divide the inspection grayscale images into different equal blocks according to a preset size; perform regional division on all equal blocks according to the grayscale distribution in the equal blocks to obtain normal areas and abnormal areas affected by illumination;

[0007] A detail recognition module is used to perform edge detection on each normal area and abnormal area to obtain edge pixels in each area, and determine the influence of detail features of each abnormal area according to the difference in the number of edge pixels in all normal areas and each abnormal area;

[0008] A distribution analysis module is used to determine the distribution concentration of abnormal areas according to the location distribution characteristics of all abnormal areas;

[0009] The enhancement module is used to determine the degree of illumination influence according to the degree of influence of the detail features and the degree of distribution concentration of each abnormal area; screen out the illumination influence area to be enhanced according to the illumination influence degree, and perform image enhancement on the illumination influence area to be enhanced.

[0010] Furthermore, the region division of all equally divided blocks according to the grayscale distribution in the equally divided blocks to obtain normal areas and abnormal areas affected by illumination includes:

[0011] Calculate the grayscale mean of the pixels in each equally divided block, and use the absolute value of the difference between the grayscale mean and the grayscale mean of all pixels in the inspection grayscale image as a growth indicator;

[0012] Performing regional growth on all the equally divided blocks according to the growth index to obtain growth clusters;

[0013] According to the growth indicators of each type of growth cluster, the growth clusters belonging to the normal area and the growth clusters belonging to the abnormal area are determined.

[0014] Furthermore, the determining of growth clusters belonging to normal areas and growth clusters belonging to abnormal areas for the equally divided growth indicators in each type of growth clusters includes:

[0015] The average value of the growth indexes of all equally divided blocks in each growth cluster is used as a judgment index, and the growth cluster is determined to be a normal area or an abnormal area according to the judgment index, wherein the judgment index of the normal area is smaller than the judgment index of the abnormal area.

[0016] Furthermore, the determining of the influence degree of the detail features of each abnormal area according to the difference in the number of edge pixels in all normal areas and each abnormal area includes:

[0017] The ratio of the number of edge pixels in all normal areas to the total number of all pixels in the normal area is used as the normal number feature ratio;

[0018] The ratio of the number of edge pixels in any abnormal region to the number of all pixels in the abnormal region is used as the abnormal number feature ratio of the corresponding abnormal region;

[0019] The absolute value of the difference between the normal quantity feature ratio and the abnormal quantity feature ratio is normalized and used as the influence degree of the detail features of the corresponding abnormal area.

[0020] Furthermore, the determining of the distribution concentration of abnormal areas according to the location distribution characteristics of all abnormal areas includes:

[0021] Determine the morphological center point of each abnormal region, and take the Euclidean distance between the center points of any two abnormal regions as the center distance between the corresponding two abnormal regions;

[0022] Determine the concentration index of the corresponding abnormal area according to the center distance between any abnormal area and a preset number of other abnormal areas closest to it;

[0023] The mean of the concentration index of all abnormal areas is normalized and used as the distribution concentration degree of the abnormal areas.

[0024] Furthermore, the determining of the concentration index of the corresponding abnormal area according to the center distance between any abnormal area and a preset number of other abnormal areas closest to it includes:

[0025] The average of the center distances between any abnormal area and the preset number of other abnormal areas closest to it is taken as the neighborhood mean;

[0026] The reciprocal of the neighborhood mean is normalized and used as a concentration indicator of the abnormal area.

[0027] Furthermore, determining the degree of illumination influence according to the degree of influence of the detail features and the degree of distribution concentration of each abnormal area includes:

[0028] The product of the detail feature influence degree and the distribution concentration degree is calculated, and normalized to obtain the illumination influence degree.

[0029] Further, the illumination impact area to be enhanced is selected according to the illumination impact degree, including:

[0030] The area where the illumination influence degree is greater than the preset influence threshold is taken as the illumination influence to be enhanced area.

[0031] Furthermore, image enhancement is performed on the area to be enhanced with respect to the illumination influence, including:

[0032] The illumination influence degree is used as an image enhancement coefficient, and a linear enhancement process is performed on the illumination influence area to be enhanced.

[0033] Furthermore, the preset size is a 5×5 pixel size.

[0034] The present invention has the following beneficial effects:

[0035] In the embodiments of the present invention, the uneven distribution of lighting influence in the image causes the optimization effect of unified image enhancement to be too poor, which in turn causes unclear image details and low image quality. By dividing all equally divided blocks into regions through grayscale distribution, normal areas and abnormal areas affected by lighting are obtained. The abnormal areas are not limited to highlight areas or backlight areas, but are areas that are significantly different from the entire image, thereby making the identification of abnormal areas more reliable and improving the scope of applicable scenarios. The edge texture is feature analyzed by the number of edge pixels to obtain the degree of influence of detail features. The degree of influence of detail features refers to the characteristics of the smaller background difference of the grayscale image of the same frame inspection. The degree of influence of detail features in abnormal areas affected by illumination is accurately determined through the difference between the detail performance in normal areas and the detail performance in abnormal areas. Then, the distribution concentration of abnormal areas is determined according to the position distribution characteristics of all abnormal areas. The degree of illumination influence is determined by combining the degree of influence of detail features and the degree of distribution concentration. The degree of illumination influence can be more suitable for complex and changeable low-altitude inspection scenes. At the same time, more accurate illumination influence areas to be enhanced are identified, and adaptive image enhancement processing is performed in combination with the degree of illumination influence to improve the optimization effect, thereby enhancing the clarity of image details and improving the quality of the enhanced image. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 A structural diagram of a low-altitude intelligent fusion rail transit integrated system based on BIM technology provided by an embodiment of the present invention;

[0038] Figure 2 A schematic diagram of an inspection grayscale image provided by an embodiment of the present invention;

[0039] Figure 3 A schematic diagram of area division provided by an embodiment of the present invention;

[0040] Figure 4 A schematic diagram of an enhanced image provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation method, structure, characteristics and effects of a low-altitude intelligent fusion rail transit integration system based on BIM technology proposed by the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0042] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0043] Some explanations of terms in this plan:

[0044] Building Information Modeling (BIM) is a process of establishing a building model based on the relevant information data of a construction project, simulating the real information of a building through digital information simulation.

[0045] Integrated rail transit uses BIM technology to construct a three-dimensional information model of rail transit and its surrounding areas, and transmits drone inspection data to the integrated system through an integrated network. The system uniformly displays the cruise route and real-time video data, and uses video algorithms to identify hazardous sources. At the same time, the system conducts a closed-loop inspection for subsequent rectification.

[0046] Drone inspection is a technology that uses unmanned aerial vehicles (UAVs) to inspect equipment and facilities. It is widely used in multiple industries, such as electricity, oil and gas, transportation, agriculture, and construction. Drone inspection can improve inspection efficiency, reduce labor costs, enhance safety, and provide higher quality data collection.

[0047] The following is a detailed description of a specific scheme of a low-altitude intelligent fusion rail transit integration system based on BIM technology provided by the present invention in conjunction with the accompanying drawings.

[0048] See also Figure 1 , which shows a structural diagram of a low-altitude intelligent fusion rail transit integrated system based on BIM technology provided by an embodiment of the present invention, including:

[0049] The acquisition module 101 is used to acquire the inspection grayscale images of different frames during the low-altitude inspection process, and divide the inspection grayscale images into different equal blocks according to the preset size; divide all the equal blocks into regions according to the grayscale distribution in the equal blocks to obtain normal areas and abnormal areas affected by lighting.

[0050] This solution can configure a high-resolution camera on the drone to shoot the area to be inspected, and transmit the captured images to the integrated system. The integrated system uniformly displays the cruise route and real-time image data, and combines video algorithms to identify the source of danger. It is understandable that drones are usually inspected at low altitudes below 100m. During the actual inspection process, images of the area to be inspected are obtained by interval shooting or recording. This facilitates the subsequent construction of the BIM model.

[0051] In the embodiment of the present invention, each frame in the image can be extracted and grayscaled to obtain an inspection grayscale image, which is convenient for subsequent image analysis. Figure 2 , Figure 2 A schematic diagram of an inspection grayscale image provided by an embodiment of the present invention.

[0052] The preset size is the size of the equally divided blocks. Optionally, in some embodiments of the present invention, the preset size may be specifically, for example, a 5×5 pixel size, that is, the inspection grayscale image is divided into a plurality of adjacent non-overlapping equally divided blocks according to a 5×5 size. Of course, in other embodiments of the present invention, the preset size may also be adjusted according to actual detection requirements, and there is no limitation to this.

[0053] In actual analysis, a lot of content needs to be calculated for each pixel point, and the analysis efficiency is too low. In addition, the area to be inspected usually does not require too detailed analysis to achieve high-quality inspection results. Therefore, this solution sets equal blocks to reduce the amount of analysis and improve efficiency.

[0054] Each equally divided block can be in an area with normal lighting effects, or an area with abnormal lighting effects. The areas with abnormal lighting effects mainly include strong light areas and backlight areas. Both the strong light areas and the backlight areas may suffer from detail loss due to lighting effects. At this time, an integrated system is required to perform image enhancement to improve image contrast and thereby improve image quality.

[0055] It is understandable that the image acquisition device also has a certain adaptive adjustment function, but this adjustment is limited to the overall enhancement of the video according to the current scene. When low-altitude inspections encounter complex scenes, for example, most areas of the overall image are in normal photosensitivity, while some areas are affected by strong light or abnormal backlight. At this time, unified image enhancement will result in poor optimization effect and still cannot solve the problem of image detail display.

[0056] Based on this, this solution identifies normal areas and abnormal areas to facilitate subsequent image enhancement of abnormal areas.

[0057] It is understandable that the grayscale value of an area affected by strong light or abnormal backlight usually presents a feature that is quite different from that of most scenes in the image, and specific analysis is performed based on this feature.

[0058] Furthermore, in some embodiments of the present invention, all equally divided blocks are divided into regions according to the grayscale distribution within the equally divided blocks to obtain normal areas and abnormal areas affected by lighting, including: calculating the grayscale mean of the pixels in each equally divided block, and taking the absolute value of the difference between the grayscale mean and the grayscale mean of all pixels in the inspection grayscale image as a growth index; performing regional growth on all equally divided blocks according to the growth index to obtain growth clusters; and determining the growth clusters belonging to the normal area and the growth clusters belonging to the abnormal area for the growth index of the equally divided blocks in each type of growth cluster.

[0059] It should be noted that, since the distribution of abnormal regions and normal regions is inconsistent, there may be multiple abnormal regions and normal regions. Through regional growing, equally divided blocks with similar performances can be aggregated into growth clusters to perform regional analysis on the growth clusters.

[0060] In the embodiment of the present invention, since the abnormal illumination areas are usually concentrated in a specific range, regional growth can be performed to perform regional division, and the growth clusters belonging to the normal area and the growth clusters belonging to the abnormal area are determined based on the growth indicators of the equal blocks in each growth cluster.

[0061] In an embodiment of the present invention, the growth index represents a numerical index for performing regional growth analysis. The grayscale mean of the pixels in the equally divided block can be used as the grayscale feature of the equally divided block, and a difference analysis is performed between the grayscale mean of the equally divided block and the grayscale mean of all the pixels in the inspection grayscale image, that is, the absolute value of the difference between the grayscale mean of the equally divided block and the grayscale mean of all the pixels in the inspection grayscale image is calculated. The larger the value of the absolute value of the difference, the greater the grayscale difference between the corresponding equally divided block and the image as a whole, that is, the more likely the equally divided block itself is to be in an abnormal grayscale area, and the greater its growth index; and the normal area is the majority of the area in the image, so the grayscale difference between it and the overall grayscale of the image is smaller, that is, the growth index is smaller; based on this, normal areas and abnormal areas can be divided.

[0062] Furthermore, in some embodiments of the present invention, growth clusters belonging to normal areas and growth clusters belonging to abnormal areas are determined based on the growth indicators of equally divided blocks in each type of growth cluster, including: taking the average of the growth indicators of all equally divided blocks in each growth cluster as a judgment indicator, and determining whether the growth cluster is a normal area or an abnormal area based on the judgment indicator, wherein the judgment indicator of the normal area is smaller than the judgment indicator of the abnormal area.

[0063] In an embodiment of the present invention, the judgment index of each growth cluster can be calculated, and then, the threshold value of the judgment index can be determined by using the Otsu method, and the growth cluster greater than the threshold value can be regarded as an abnormal area, and the growth cluster less than or equal to the threshold value can be regarded as a normal area.

[0064] See also Figure 2 and Figure 3 , Figure 3 This is a schematic diagram of region division provided by an embodiment of the present invention. In the embodiment of the present invention, cluster analysis can be performed according to the judgment index to obtain normal regions and abnormal regions. In the embodiment of the present invention, k-means clustering can be used, and the k value is set to 2, thereby realizing region division to obtain normal regions and abnormal regions.

[0065] Of course, the threshold value of the judgment index may also be directly set, for example, 0.8, and the growth clusters greater than 0.8 are regarded as abnormal areas, and the growth clusters less than or equal to 0.8 are regarded as normal areas. There is no restriction on this.

[0066] The detail recognition module 102 is used to perform edge detection on each normal area and abnormal area to obtain edge pixels in each area, and determine the impact degree of detail features of each abnormal area based on the difference in the number of edge pixels in all normal areas and each abnormal area.

[0067] After determining the normal area and the abnormal area, the details of the normal area and the abnormal area can be specifically analyzed. It is understandable that under normal lighting conditions, the details are relatively stable, but when the lighting is abnormal, the details are affected by the lighting due to both the highlight and the backlight, resulting in detail loss. The features and textures of these areas may become blurred due to changes in lighting, making it impossible to fully display the originally clear details, making it difficult to obtain comprehensive information in these areas. Therefore, in order to analyze the detail loss phenomenon, the degree of influence of detail features is introduced in the embodiment of the present invention.

[0068] In the embodiment of the present invention, edge information is used as a representation of details, and the influence degree of detail features is determined by analyzing the edge information.

[0069] Among them, edge detection can be specifically, for example, Canny edge detection, which determines the edge pixels in each normal area or abnormal area. It can be understood that the more edge pixels there are, the more complex the texture in the corresponding area is, that is, the more perfect the detail features are retained. However, due to the influence of abnormal lighting, details may be lost, resulting in a smaller number of edge pixels.

[0070] Furthermore, in some embodiments of the present invention, the degree of influence of detail features of each abnormal area is determined based on the difference in the proportion of the number of edge pixels in all normal areas and each abnormal area, including: taking the ratio of the number of edge pixels in all normal areas to the total number of all pixels in the normal areas as the normal quantity feature ratio; taking the ratio of the number of edge pixels in any abnormal area to the number of all pixels in the abnormal area as the abnormal quantity feature ratio of the corresponding abnormal area; and normalizing the absolute value of the difference between the normal quantity feature ratio and the abnormal quantity feature ratio as the degree of influence of detail features of the corresponding abnormal area.

[0071] In the embodiment of the present invention, by determining the normal number feature ratio and the abnormal number feature ratio, the absolute value of the difference between the normal number feature ratio and the abnormal number feature ratio is calculated. The larger the value is, the greater the difference between the detail texture in the abnormal area and the detail texture in the normal area is. In other words, the larger the numerical value of the influence degree of the detail features in the abnormal area is, the less texture details in the abnormal area that are affected by the light are retained.

[0072] The distribution analysis module 103 is used to determine the distribution concentration degree of the abnormal regions according to the position distribution characteristics of all abnormal regions.

[0073] Among them, the distribution concentration degree is mainly analyzed for the distribution characteristics of the abnormal area itself. It can be understood that the influence of light usually has a certain concentration characteristic, that is, if a certain point is affected by strong light, then its adjacent points are more likely to be affected by strong light, that is, the corresponding regional concentration characteristics are presented. Similarly, the more concentrated the distribution of areas affected by abnormal light in a certain area, the more serious the abnormal situation of the abnormal area affected by light. Therefore, the distribution concentration degree is used for specific analysis in the embodiment of the present invention.

[0074] Furthermore, in some embodiments of the present invention, the distribution concentration degree of abnormal areas is determined based on the position distribution characteristics of all abnormal areas, including: determining the morphological center point of each abnormal area, and taking the Euclidean distance between the center points of any two abnormal areas as the center distance between the two corresponding abnormal areas; determining the concentration index of the corresponding abnormal area based on the center distance between any abnormal area and a preset number of other abnormal areas that are closest to it; and normalizing the mean of the concentration indexes of all abnormal areas as the distribution concentration degree of the abnormal areas.

[0075] Among them, the location information of each abnormal area is characterized by determining the morphological center point, and the distribution concentration degree is analyzed by the Euclidean distance between the center points. The smaller the center distance, the more concentrated the distribution. Therefore, the concentration index of the corresponding abnormal area can be determined according to the center distance.

[0076] The preset number is the number of other adjacent abnormal areas for centralized analysis. The preset number in the embodiment of the present invention may be specifically 5, for example, or may be adjusted according to actual needs, without limitation.

[0077] Furthermore, in some embodiments of the present invention, a concentration index of the corresponding abnormal area is determined based on the center distance between any abnormal area and a preset number of other abnormal areas that are closest to it, including: taking the average of the center distances between any abnormal area and a preset number of other abnormal areas that are closest to it as the neighborhood mean; and normalizing the inverse of the neighborhood mean as the concentration index of the abnormal area.

[0078] Among them, the neighborhood mean is obtained by calculating the mean of the center distance. The larger the value of the neighborhood mean, the larger the center distance, that is, the more dispersed the distribution, and the smaller the value of the neighborhood mean, the smaller the center distance, that is, the more concentrated the distribution. Therefore, by calculating the inverse of the neighborhood mean, normalization is used as a concentration indicator of the abnormal area.

[0079] In the embodiment of the present invention, the concentration indexes of all abnormal regions may be analyzed in an integrated manner, that is, the mean of the concentration indexes of all abnormal regions may be calculated and normalized to be the distribution concentration degree of the abnormal regions.

[0080] The distribution concentration degree can characterize the concentration of distributions of different abnormal regions. It should be noted that when there is only one abnormal region, the corresponding distribution concentration degree is 1.

[0081] The enhancement module 104 is used to determine the degree of illumination influence according to the degree of influence of detail features and the degree of distribution concentration of each abnormal area; select the illumination influence area to be enhanced according to the illumination influence degree, and perform image enhancement on the illumination influence area to be enhanced.

[0082] In the embodiment of the present invention, combined with the above analysis, the degree of influence of detail features represents the abnormal situation in which details are affected by light. The larger the value of the degree of influence of detail features in the abnormal area, the less texture details are retained in the abnormal area affected by light, and the corresponding abnormal area needs image enhancement processing.

[0083] Among them, the greater the distribution concentration, the more concentrated the abnormal area distribution is, that is, the more concentrated the area affected by light in the inspection grayscale image of the frame. In order to distinguish the concentrated area more effectively, image enhancement is more needed to improve the image recognition.

[0084] In summary, in the embodiments of the present invention, the degree of illumination influence is calculated to characterize the influence of illumination on the corresponding abnormal area. The larger the value of the degree of illumination influence, the less texture details of the corresponding abnormal area are retained and the distribution is more concentrated.

[0085] Furthermore, in some embodiments of the present invention, the degree of illumination influence is determined according to the degree of detail feature influence and the degree of distribution concentration of each abnormal area, including: calculating the product of the degree of detail feature influence and the degree of distribution concentration, and normalizing to obtain the degree of illumination influence.

[0086] In the embodiment of the present invention, combined with the above analysis, the degree of influence of detail features represents the abnormal situation in which details are affected by light. The larger the value of the degree of influence of detail features in the abnormal area, the less texture details are retained in the abnormal area affected by light, and the corresponding abnormal area needs image enhancement processing.

[0087] Among them, the greater the distribution concentration, the more concentrated the abnormal area distribution is, that is, the more concentrated the area affected by light in the inspection grayscale image of the frame. In order to distinguish the concentrated area more effectively, image enhancement is more needed to improve the image recognition.

[0088] In summary, in the embodiments of the present invention, the degree of illumination influence is calculated to characterize the influence of illumination on the corresponding abnormal area. The larger the value of the degree of illumination influence, the less texture details of the corresponding abnormal area are retained and the distribution is more concentrated.

[0089] Therefore, the product of the influence degree of detail features and the distribution concentration degree is calculated, and the maximum and minimum values ​​can be linearly normalized to obtain the illumination influence degree.

[0090] Since the greater the degree of illumination influence, the greater the corresponding abnormal area is affected by illumination, and the more blurred the texture details in the corresponding abnormal area are and the more concentrated the distribution is, therefore, it is necessary to improve the degree of image enhancement. Therefore, in an embodiment of the present invention, the abnormal area is further screened to obtain the illumination influence area to be enhanced that needs to be image enhanced.

[0091] Furthermore, in some embodiments of the present invention, the lighting impact area to be enhanced is screened out according to the lighting impact degree, including: taking the area with a lighting impact degree greater than a preset impact threshold as the lighting impact area to be enhanced.

[0092] The preset impact threshold is a threshold value of the illumination impact degree. In the embodiment of the present invention, the preset impact threshold can be set to 0.5, that is, when the illumination impact degree is greater than 0.5, it is used as an illumination impact area to be enhanced.

[0093] After determining the area to be enhanced due to the illumination influence, it is necessary to perform image enhancement on the area to be enhanced due to the illumination influence. In the embodiment of the present invention, in order to achieve adaptive image enhancement, the illumination influence degree can be directly used as the image enhancement coefficient, and linear enhancement processing can be performed on the area to be enhanced due to the illumination influence. Of course, in other embodiments of the present invention, a standard enhancement coefficient can also be preset, and then the standard enhancement coefficient is adjusted according to the illumination influence degree to obtain an adaptive enhancement coefficient for each area to be enhanced due to the illumination influence. The adjustment is all linear adjustment, which is set according to the actual situation. Image enhancement processing is also a technology well known to those skilled in the art, and no further limitation or elaboration is made thereon.

[0094] See also Figure 2 and Figure 4 , Figure 4 A schematic diagram of an enhanced image provided by an embodiment of the present invention, compared with Figure 2 and Figure 4 ,After image enhancement, the texture becomes clearer and the overall image quality is improved.

[0095] Therefore, each illumination-affected area to be enhanced in each frame of the inspection grayscale image can be adaptively enhanced, thereby improving the overall image quality and avoiding illumination effects. At the same time, all enhanced inspection grayscale images are synthesized using image editing software in time sequence, thereby obtaining an enhanced image of the inspection area, which is used for more accurate and reliable low-altitude intelligent inspections.

[0096] In the embodiments of the present invention, the uneven distribution of lighting influence in the image causes the optimization effect of unified image enhancement to be too poor, which in turn causes unclear image details and low image quality. By dividing all equally divided blocks into regions through grayscale distribution, normal areas and abnormal areas affected by lighting are obtained. The abnormal areas are not limited to highlight areas or backlight areas, but are areas that are significantly different from the entire image, thereby making the identification of abnormal areas more reliable and improving the scope of applicable scenarios. The edge texture is feature analyzed by the number of edge pixels to obtain the degree of influence of detail features. The degree of influence of detail features refers to the characteristics of the smaller background difference of the grayscale image of the same frame inspection. The degree of influence of detail features in abnormal areas affected by illumination is accurately determined through the difference between the detail performance in normal areas and the detail performance in abnormal areas. Then, the distribution concentration of abnormal areas is determined according to the position distribution characteristics of all abnormal areas. The degree of illumination influence is determined by combining the degree of influence of detail features and the degree of distribution concentration. The degree of illumination influence can be more suitable for complex and changeable low-altitude inspection scenes. At the same time, more accurate illumination influence areas to be enhanced are identified, and adaptive image enhancement processing is performed in combination with the degree of illumination influence to improve the optimization effect, thereby enhancing the clarity of image details and improving the quality of the enhanced image.

[0097] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0098] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A low-altitude intelligent integrated rail transit system based on BIM technology, characterized in that: include: An acquisition module is used to acquire inspection grayscale images of different frames during the low-altitude inspection process, and divide the inspection grayscale images into adjacent non-overlapping equal blocks according to a preset size; All equally divided blocks are divided into regions according to the grayscale distribution in the equally divided blocks to obtain normal areas and abnormal areas affected by lighting; A detail recognition module is used to perform edge detection on each normal area and abnormal area to obtain edge pixels in each area, and determine the influence of detail features of each abnormal area according to the difference in the number of edge pixels in all normal areas and each abnormal area; A distribution analysis module is used to determine the distribution concentration of abnormal areas according to the location distribution characteristics of all abnormal areas; An enhancement module, used to determine the degree of illumination influence according to the degree of influence of the detail features and the degree of distribution concentration of each abnormal area; screen out the illumination influence area to be enhanced according to the illumination influence degree, and perform image enhancement on the illumination influence area to be enhanced; The method of dividing all equally divided blocks into regions according to the grayscale distribution in the equally divided blocks to obtain normal regions and abnormal regions affected by illumination includes: Calculate the grayscale mean of the pixels in each equally divided block, and use the absolute value of the difference between the grayscale mean and the grayscale mean of all pixels in the inspection grayscale image as a growth indicator; Performing regional growth on all the equally divided blocks according to the growth index to obtain growth clusters; According to the growth index of each type of growth cluster, the growth clusters belonging to the normal area and the growth clusters belonging to the abnormal area are determined; Determining the influence of the detail features of each abnormal area according to the difference in the number of edge pixels in all normal areas and each abnormal area includes: The ratio of the number of edge pixels in all normal areas to the total number of all pixels in the normal area is used as the normal number feature ratio; The ratio of the number of edge pixels in any abnormal region to the number of all pixels in the abnormal region is used as the abnormal number feature ratio of the corresponding abnormal region; The absolute value of the difference between the normal quantity feature ratio and the abnormal quantity feature ratio is normalized and used as the influence degree of the detail features of the corresponding abnormal area.

2. A low-altitude intelligent integrated rail transit system based on BIM technology as claimed in claim 1, characterized in that: The step of determining the growth clusters belonging to the normal area and the growth clusters belonging to the abnormal area for the growth indicators of each type of growth cluster includes: The average value of the growth indexes of all equally divided blocks in each growth cluster is used as a judgment index, and the growth cluster is determined to be a normal area or an abnormal area according to the judgment index, wherein the judgment index of the normal area is smaller than the judgment index of the abnormal area.

3. The low-altitude intelligent integrated rail transit system based on BIM technology as claimed in claim 1 is characterized in that: Determining the distribution concentration of abnormal areas according to the location distribution characteristics of all abnormal areas includes: Determine the morphological center point of each abnormal region, and take the Euclidean distance between the center points of any two abnormal regions as the center distance between the corresponding two abnormal regions; Determine the concentration index of the corresponding abnormal area according to the center distance between any abnormal area and a preset number of other abnormal areas closest to it; The mean of the concentration index of all abnormal areas is normalized and used as the distribution concentration degree of the abnormal areas.

4. A low-altitude intelligent integrated rail transit system based on BIM technology as claimed in claim 3, characterized in that: The step of determining the concentration index of the corresponding abnormal area according to the center distance between any abnormal area and a preset number of other abnormal areas closest to it includes: The average of the center distances between any abnormal area and the preset number of other abnormal areas closest to it is taken as the neighborhood mean; The reciprocal of the neighborhood mean is normalized and used as a concentration indicator of the abnormal area.

5. The low-altitude intelligent integrated rail transit system based on BIM technology as claimed in claim 1 is characterized in that: The step of determining the degree of illumination influence according to the degree of influence of the detail features and the degree of distribution concentration of each abnormal area includes: The product of the detail feature influence degree and the distribution concentration degree is calculated, and normalized to obtain the illumination influence degree.

6. The low-altitude intelligent integrated rail transit system based on BIM technology as claimed in claim 1 is characterized in that: According to the degree of illumination influence, the illumination influence area to be enhanced is selected, including: The area where the illumination influence degree is greater than the preset influence threshold is taken as the illumination influence to be enhanced area.

7. The low-altitude intelligent integrated rail transit system based on BIM technology as claimed in claim 1 is characterized in that: Performing image enhancement on the area to be enhanced due to the illumination influence, including: The illumination influence degree is used as an image enhancement coefficient, and a linear enhancement process is performed on the illumination influence area to be enhanced.

8. The low-altitude intelligent integrated rail transit system based on BIM technology as claimed in claim 1 is characterized in that: The preset size is 5×5 pixels.

Images