A construction progress assessment method based on construction images

Abstract

This invention discloses a construction progress assessment method based on construction images, applicable to the construction field. The method includes the following steps: equipping a drone with a wireless communication module for a meteorological station; acquiring meteorological parameter data and meteorological images; developing flight plans for abnormal weather conditions based on comprehensive weather forecasts; improving the quality of abnormal construction images using dark channel defogging and extracting key information about construction progress using a SegNet deep network; comparing the key construction progress information with actual progress information, planned progress information, and historical progress information from the previous period; and visually presenting the construction progress assessment results based on the construction images through 3D modeling. This invention establishes a time series model and extracts key features to comprehensively predict weather conditions for the coming week, enabling advance weather planning and the implementation of corresponding measures to address potential abnormal weather.

Description

Technical Field

[0001] This invention relates to the field of building construction, and more particularly to a method for evaluating building progress based on construction images. Background Technology

[0002] The construction industry is a cornerstone of the national economy, playing an indispensable role in economic and social progress. It is widely believed that precise supervision of construction is a key element in ensuring the success of construction projects, and the core of this supervision is the accurate assessment of construction progress. The construction phase is a critical period in the life cycle of a construction project, with the majority of time and funds invested in this stage. Therefore, progress assessment and management during the construction phase are essential for achieving project objectives.

[0003] Assessing construction progress is an ongoing process that requires the continuous collection and updating of on-site data. This results in a massive flow of information, which continues to increase as the project progresses. The dynamism and sheer volume of this real-time information present challenges for its storage, sharing, and management. Traditional methods primarily rely on manual collection of on-site data—through manual measurements and written records—followed by processing the data and generating reports. This traditional data collection approach is no longer adequate for the needs of modern construction projects, and its limitations are becoming increasingly apparent.

[0004] In the modern construction industry, drone technology has become an effective tool for assisting in the assessment and monitoring of construction progress. Drones offer many advantages, particularly in data collection and monitoring, but they also have some potential limitations.

[0005] First, the use of drones offers significant advantages in construction. Drones can cover a wide construction area in a short time, capturing high-resolution images and video data. This data can be used to generate 3D models of the construction site, helping supervisors and management teams better understand the actual progress of the project. Furthermore, drones can fly regularly, collecting data in real time, thus providing accurate information about construction progress and site conditions.

[0006] However, drone technology also has some limitations, one of which is its susceptibility to adverse weather and environmental conditions. Severe weather conditions, such as strong winds, heavy rain, or dense fog, may limit the drone's flight capabilities, thereby affecting the frequency and quality of data collection. This may lead to inaccurate and discontinuous data, thus reducing the ability to accurately assess construction progress.

[0007] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0008] To overcome the above problems, this invention aims to propose a construction progress assessment method based on construction images, which addresses the issue that severe weather conditions, such as strong winds, heavy rain, or dense fog, may limit the flight capabilities of drones, thereby affecting the frequency and quality of data collection.

[0009] Therefore, the specific technical solution adopted by the present invention is as follows:

[0010] A construction progress assessment method based on construction images includes the following steps:

[0011] S1. Set up a meteorological station at the construction site and equip it with a drone and a wireless communication module for the meteorological station;

[0012] S2. The meteorological station predicts the weather conditions at the construction site for the next week, obtains meteorological parameter data and meteorological images, makes a comprehensive forecast of the future weather based on the meteorological parameter data and meteorological images, and classifies the weather conditions into normal weather and abnormal weather based on the comprehensive forecast results.

[0013] S3. Develop flight plans for abnormal weather conditions based on comprehensive weather forecasts and acquire images of abnormal building construction.

[0014] S4. Improve the quality of abnormal building construction images by using dark channel dehazing, and extract key information on construction progress from the enhanced abnormal building construction images using the SegNet deep network.

[0015] S5. Compare key construction progress information with actual progress information, planned progress information, and historical progress information from the previous period to comprehensively assess progress deviations.

[0016] S6. Visualize the construction progress assessment results based on construction images through 3D modeling, and generate a progress assessment report containing images, data analysis and recommendations.

[0017] Optionally, the meteorological station predicts the weather conditions at the construction site for the next week, acquires meteorological parameter data and meteorological images, makes a comprehensive forecast of the future weather based on the meteorological parameter data and meteorological images, and classifies the weather conditions into normal weather and abnormal weather based on the comprehensive forecast results, including the following steps:

[0018] S21. The meteorological station collects meteorological parameter data, which includes at least temperature data, humidity data, air pressure data, wind speed data, and wind direction data.

[0019] S22. Meteorological stations collect meteorological images, which include at least cloud images, radar images, and infrared images;

[0020] S23. Extract key features from meteorological parameter data and meteorological images;

[0021] S24. Establish a time series model, input meteorological parameter data and key features of meteorological images, and make a comprehensive forecast of future weather to obtain the weather conditions for the next week.

[0022] S25. Based on the predicted weather conditions, set thresholds to distinguish between normal weather and abnormal weather;

[0023] S26. If the weather conditions are within the set threshold range, they are determined to be normal weather.

[0024] If the weather exceeds the set threshold, it is considered abnormal weather.

[0025] Optionally, the meteorological station forecasts the weather conditions at the construction site for the next week, acquires meteorological parameter data and meteorological images, makes a comprehensive forecast of the future weather based on the meteorological parameter data and meteorological images, and classifies the weather conditions into normal weather and abnormal weather based on the comprehensive forecast results, which also includes the following steps:

[0026] If the overall forecast indicates normal weather, then a flight plan for normal weather conditions will be formulated.

[0027] Acquire construction images of the target building and extract key information about the construction progress.

[0028] Optionally, the step of improving the quality of abnormal building construction images by using dark channel dehazing and extracting key information about construction progress from the enhanced abnormal building construction images using a SegNet deep network includes the following steps:

[0029] S41. Obtain the original high-resolution abnormal building construction image from the abnormal building construction image;

[0030] S42. Downsample the original high-resolution abnormal building construction image to obtain a downsampled image with reduced resolution;

[0031] S43. Calculate the dark channel on the downsampled image, establish a minimum value table to accelerate the calculation, and estimate the preliminary transmittance of the downsampled image based on the dark channel results.

[0032] S44. Perform morphological edge detection on the original high-resolution abnormal building construction image to obtain an edge binary map, and use the edge binary map to optimize the initial transmittance of the abnormal building construction image to obtain the optimized transmittance of the abnormal building construction image.

[0033] S45. Measure the ambient light intensity, and calculate the preliminary dehazing image based on the ambient light, the transmittance of the optimized abnormal building construction image, and the original high-resolution abnormal building construction image using the dehazing model.

[0034] S46. Upsample the initial dehazed image to obtain a dehazed image of the same size as the original high-resolution abnormal building construction image, and process it to obtain the final dehazed image;

[0035] S47. Collect a dataset of abnormal building construction images containing different building stages, and annotate key building element data;

[0036] S48. Train a SegNet network model using labeled key building element data.

[0037] S49. Use the trained SegNet network model to predict the final dehazed image and obtain the semantic segmentation results of the building elements.

[0038] S410. Based on the presence of different building elements in the segmentation results, assess the current construction progress information.

[0039] Optionally, the step of calculating the dark channel on the downsampled image, establishing a minimum value table to accelerate the calculation, and estimating the preliminary transmittance of the downsampled image based on the dark channel results includes the following steps:

[0040] S431. Normalize the downsampled image and map it to a set range (0-1);

[0041] S432. Slide a window in the downsampled image, and in each window, traverse all pixels to find the minimum value of the red channel, green channel, and blue channel;

[0042] S433. Take the minimum value of the red, green and blue channels of each window as the dark channel value of that window;

[0043] S434. Combine the dark channel values ​​of the red, green, and blue channels to obtain the dark channel map of the downsampled image;

[0044] S435. Select the pixel with the lowest empirical value (usually 0.1%) from the dark channel image;

[0045] S436. Find the brightness value corresponding to the empirical value pixel in the original downsampled image, take the average value as the global atmospheric light value, and calculate the preliminary transmittance of the downsampled image.

[0046] The formula for calculating the initial transmittance of the downsampled image is as follows:

[0047]

[0048] In the formula, ω Indicates empirical weights;

[0049] J represents the dark channel value;

[0050] A represents the global atmospheric value;

[0051] K represents the initial transmittance of the downsampled image.

[0052] Optionally, the step of performing morphological edge detection on the original high-resolution abnormal building construction image to obtain an edge binary map, and then using the edge binary map to optimize the initial transmittance of the abnormal building construction image to obtain the optimized transmittance of the abnormal building construction image includes the following steps:

[0053] S441. Perform high-pass filtering on the original high-resolution abnormal building construction image to filter out uniform areas and retain edges and details;

[0054] S442. Perform Sobel edge detection on the filtered image to detect image edges;

[0055] S443. Perform double thresholding on the Sobel edge detection results to obtain a binary edge map;

[0056] S444: Perform dilation and erosion operations on the edge binary image to remove noise;

[0057] S445. Use the initial transmittance of the downsampled image as the initial transmittance of the abnormal building construction image.

[0058] S446. For each pixel in a non-edge region, calculate the weight based on the spatial distance between it and neighboring pixels in a preset neighborhood, and calculate the weighted average of the transmittance in that neighborhood. Set the transmittance of the pixel to the calculated weighted average transmittance to adjust the transmittance of the non-edge region.

[0059] S447. Repeat step S446 until the transmittance of all non-edge regions is smoothed to the average value within the neighborhood, and the optimized original image transmittance is obtained.

[0060] Optionally, the step of calculating a weight for each pixel in a non-edge region based on its spatial distance to neighboring pixels within a preset neighborhood, calculating a weighted average of the transmittance within that neighborhood, and setting the transmittance of the pixel to the calculated weighted average transmittance to adjust the transmittance of the non-edge region includes the following steps:

[0061] S4461. Define the neighborhood size. For each pixel in a non-edge region, traverse each pixel within its defined neighborhood.

[0062] S4462. Calculate the spatial distance between a pixel and each neighboring pixel using Euclidean distance.

[0063] S4463. Set the influence weight of each pixel, where the weight is inversely proportional to the distance;

[0064] S4464. For each neighboring pixel, multiply the transmittance by the corresponding weight, then sum all the products to get the weighted transmittance. Next, calculate the sum of all weight values, divide the weighted transmittance by the weight sum, and calculate the weighted average transmittance of the pixels in the neighborhood.

[0065] S4465. Set the transmittance of the pixel to the weighted average transmittance obtained in the previous step.

[0066] The formula for calculating the spatial distance between a pixel and each pixel in its neighborhood using Euclidean distance is as follows:

[0067]

[0068] In the formula, (x1, y1) represents the coordinates of the current pixel.

[0069] (x2, y2) represents the coordinates of a pixel in the neighborhood.

[0070] Optionally, the step of measuring ambient light intensity and calculating a preliminary dehazed image using a dehazing model based on ambient light, the optimized transmittance of the abnormal building construction image, and the original high-resolution abnormal building construction image includes the following steps:

[0071] S451. In the original high-resolution abnormal building construction image, select the brightest empirical value pixel in the image.

[0072] S452. Calculate the average value of the red, green, and blue channels in the brightest empirical value pixel as the global ambient light intensity;

[0073] S453. According to the dehazing formula, traverse all pixels. For each pixel, subtract the ambient light intensity from the color value of the pixel in the red, green and blue channels to remove the influence of ambient light on the image.

[0074] S454. Add the ambient light intensity back to the red, green, and blue channels to restore the color values ​​of the pixels after dehazing, and obtain a preliminary dehazed image.

[0075] Optionally, the step of using the trained SegNet network model to predict the final dehazed image and obtain the semantic segmentation result of the building elements includes the following steps:

[0076] S491. Normalize the final dehazed image after dehazing.

[0077] S492. Input the processed final dehazed image into the trained SegNet model for forward computation;

[0078] S493. The encoder of the SegNet model extracts the features of the image, and the decoder upsamples the feature map to the original image size.

[0079] S494. While upsampling the decoder, the feature map corresponding to the encoder is passed to the decoder through a skip connection. Then, the output feature map is classified by softmax to obtain the probability of each pixel belonging to each category.

[0080] S495. Based on the category with the highest probability, classify each pixel to generate semantic segmentation results for building elements;

[0081] S496. Based on the segmentation results, calculate the existence ratio and distribution information of various architectural elements;

[0082] S497. Visualize and compare the segmentation results with the original image, and check the segmentation effect.

[0083] Optionally, assessing the current construction progress information based on the presence of different building elements in the segmentation results includes the following steps:

[0084] S4101. Determine the correspondence between different architectural elements and architectural stages;

[0085] S4102. Information on the proportion of different building elements and their area ratio in the statistical segmentation results;

[0086] S4103. Based on the statistical results, make a preliminary judgment on the current construction stage of the building;

[0087] S4104. Compare the current construction stage with the planned stage. If the current stage is behind the planned stage, then there is a delay in the progress.

[0088] If the current stage is ahead of the planned stage, there is an error in the schedule, and the quality needs to be rechecked;

[0089] S4105. Determine the current progress percentage based on the completion status of different building elements, and provide an assessment result of the current construction progress by comprehensively considering factors such as building elements, corresponding project volume proportions, and quality status.

[0090] S4106. Compare the evaluation results with historical data and planned progress to determine the progress deviation.

[0091] Compared with the prior art, this application has the following advantages:

[0092] 1. This invention, by setting up meteorological measurement stations and drones, can collect meteorological parameter data and meteorological images in real time to monitor the weather conditions at construction sites. This allows for timely acquisition of the latest weather information, providing decision-making support for site managers. By utilizing meteorological parameter data and images, and establishing time series models and extracting key features, a comprehensive forecast of the weather conditions for the coming week can be made. This enables advance weather planning and the implementation of corresponding measures to address potential abnormal weather. By achieving real-time wireless communication between the meteorological measurement stations and the drone control stations, drone flight plans under normal weather conditions can be formulated based on the comprehensive forecast results, improving monitoring efficiency and reducing the waste of human resources.

[0093] 2. This invention effectively improves the clarity and contrast of abnormal building construction images through dark channel dehazing technology, making the details in the images more obvious and providing a better foundation for subsequent analysis and processing. By using the deep learning network SegNet, key building elements can be accurately extracted from the dehazed images, providing accurate data support for building progress assessment based on construction images.

[0094] 3. This invention is not only applicable to conventional building construction images, but can also process images under abnormal conditions, such as images under weather conditions like fog, rain, and mist. Through morphological edge detection and transmittance optimization, it further improves the image quality, making the building elements in the image more distinct. It not only enhances the image quality, but also extracts key construction progress information from the image, providing strong data support for the management and monitoring of building construction.

[0095] 4. This invention can accurately identify and analyze key information in construction images, providing more precise image-based assessments of construction progress. Through automated image processing and analysis, it can acquire construction progress information in real time, promptly identify deviations and problems. Compared with traditional manual assessments, this invention is more objective and accurate, reducing interference from human factors. It can adjust and optimize assessment methods according to different construction projects and conditions to meet different needs. By comparing with historical data, it can predict future construction progress and potential problems, providing forward-looking suggestions for construction management. Attached Figure Description

[0096] The above-mentioned features, characteristics, and advantages of the present invention, as well as their implementation methods, will become clearer and more readily understood in conjunction with the following description of the embodiments, which are illustrated in detail with reference to the accompanying drawings. Schematic diagrams are shown here:

[0097] Figure 1 This is a flowchart of a construction progress assessment method based on construction images according to an embodiment of the present invention. Detailed Implementation

[0098] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0099] According to an embodiment of the present invention, a method for assessing construction progress based on construction images is provided.

[0100] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, the construction progress assessment method based on construction images according to an embodiment of the present invention includes the following steps:

[0101] S1. Set up a meteorological station at the construction site and equip it with a drone and a wireless communication module for the meteorological station.

[0102] It should be explained that the site selection for a meteorological station needs to consider the following factors: distance from obstructions, good ventilation, and freedom from construction site interference. It is typically located in an open, high-altitude location. The station needs to be equipped with meteorological measuring instruments such as thermometers, hygrometers, anemometers, precipitation meters, and sunshine meters. It also needs to be equipped with data acquisition modules and transmission equipment. A power supply system for the meteorological station is also required, typically including solar panels and storage batteries. Wireless communication modules such as GPRS and CDMA, or self-organizing wireless sensor networks, are used to ensure real-time wireless communication with the drone control station. The meteorological station needs to have certain computing power to store, process, and transmit meteorological data. This is usually achieved using embedded computers or industrial control computers. The meteorological station can also integrate video monitoring functions, acquiring images through meteorological cameras to assist in weather condition assessment. Operation and maintenance procedures for the meteorological station need to be established, including instrument calibration, data maintenance, and equipment repair and maintenance.

[0103] S2. The meteorological station predicts the weather conditions at the construction site for the next week, obtains meteorological parameter data and meteorological images, makes a comprehensive forecast of the future weather based on the meteorological parameter data and meteorological images, and classifies the weather conditions into normal weather and abnormal weather based on the comprehensive forecast results.

[0104] Preferably, the meteorological station predicts the weather conditions at the construction site for the next week, acquires meteorological parameter data and meteorological images, makes a comprehensive forecast of the future weather based on the meteorological parameter data and meteorological images, and classifies the weather conditions into normal weather and abnormal weather based on the comprehensive forecast results, including the following steps:

[0105] S21. The meteorological station collects meteorological parameter data, which includes at least temperature data, humidity data, air pressure data, wind speed data, and wind direction data.

[0106] S22. Meteorological stations collect meteorological images, which include at least cloud images, radar images, and infrared images;

[0107] S23. Extract key features from meteorological parameter data and meteorological images;

[0108] S24. Establish a time series model, input meteorological parameter data and key features of meteorological images, and make a comprehensive forecast of future weather to obtain the weather conditions for the next week.

[0109] S25. Based on the predicted weather conditions, set thresholds to distinguish between normal weather and abnormal weather;

[0110] S26. If the weather conditions are within the set threshold range, they are determined to be normal weather.

[0111] If the weather exceeds the set threshold, it is considered abnormal weather.

[0112] Preferably, the meteorological station forecasts the weather conditions at the construction site for the next week, acquires meteorological parameter data and meteorological images, makes a comprehensive forecast of the future weather based on the meteorological parameter data and meteorological images, and classifies the weather conditions into normal weather and abnormal weather based on the comprehensive forecast results, further including the following steps:

[0113] If the overall forecast indicates normal weather, then a flight plan for normal weather conditions will be formulated.

[0114] Acquire construction images of the target building and extract key information about the construction progress.

[0115] It should be explained that meteorological parameter data, through time series analysis, can be used to build mathematical models to predict trends over the next week; meteorological images, through deep learning models, can be used for image classification and prediction to determine future weather conditions; fusing parameter data predictions and image predictions from multiple sources can improve prediction accuracy; after predicting weather conditions, reasonable thresholds need to be set to accurately identify abnormal weather that may affect flight; when formulating flight plans under normal weather conditions, factors such as flight route, flight altitude, and camera parameters must be considered; when formulating flight plans under abnormal weather conditions, factors such as reducing flight altitude and increasing flight stability must be considered; the flight control system can receive meteorological forecast results and automatically plan and generate optimal flight plans under different weather conditions; it can monitor meteorological changes in real time during flight and update flight plans when necessary to ensure flight safety; after flight, image processing is performed, and key construction progress information is extracted using technologies such as deep learning.

[0116] S3. Develop flight plans for abnormal weather conditions based on comprehensive weather forecasts and acquire images of abnormal building construction.

[0117] It should be clarified that abnormal weather mainly refers to meteorological conditions such as fog, haze, strong winds, and sandstorms that severely affect visibility and flight control. When developing flight plans for abnormal weather, it is necessary to consider reducing flight altitude, generally maintaining it above the minimum safe altitude, reducing flight speed, increasing flight stability, and preventing abnormal weather from affecting flight control. Shorten or adjust flight routes to avoid areas with strong winds. Install navigation aids such as barometers and radar altimeters to enhance flight safety. Preferably equip the drone with anti-fog goggles and infrared cameras to improve visibility. Increase the spacing between drone waypoints to ensure safe intervals. Monitor the drone's power, signal, and other statuses, and return to base promptly when the index approaches the limit. Monitor weather changes in real time during flight and update the flight plan as necessary. Images acquired under abnormal weather conditions require enhancement processing such as defogging and noise reduction.

[0118] S4. Improve the quality of abnormal building construction images by using dark channel dehazing, and extract key information on construction progress from the enhanced abnormal building construction images using the SegNet deep network.

[0119] Preferably, the method of improving the quality of abnormal building construction images by using dark channel dehazing and extracting key information about construction progress from the enhanced abnormal building construction images using a SegNet deep network includes the following steps:

[0120] S41. Obtain the original high-resolution abnormal building construction image from the abnormal building construction image;

[0121] S42. Downsample the original high-resolution abnormal building construction image to obtain a downsampled image with reduced resolution;

[0122] S43. Calculate the dark channel on the downsampled image, establish a minimum value table to accelerate the calculation, and estimate the preliminary transmittance of the downsampled image based on the dark channel results.

[0123] S44. Perform morphological edge detection on the original high-resolution abnormal building construction image to obtain an edge binary map, and use the edge binary map to optimize the initial transmittance of the abnormal building construction image to obtain the optimized transmittance of the abnormal building construction image.

[0124] S45. Measure the ambient light intensity, and calculate the preliminary dehazing image based on the ambient light, the transmittance of the optimized abnormal building construction image, and the original high-resolution abnormal building construction image using the dehazing model.

[0125] S46. Upsample the initial dehazed image to obtain a dehazed image of the same size as the original high-resolution abnormal building construction image, and process it to obtain the final dehazed image;

[0126] S47. Collect a dataset of abnormal building construction images containing different building stages, and annotate key building element data;

[0127] S48. Train a SegNet network model using labeled key building element data.

[0128] S49. Use the trained SegNet network model to predict the final dehazed image and obtain the semantic segmentation results of the building elements.

[0129] S410. Based on the presence of different building elements in the segmentation results, assess the current construction progress information.

[0130] In addition, the examples are summarized below:

[0131] Suppose we have an abnormal construction image, which is blurry due to fog, dust, or other factors. To improve the image quality and extract key information about the construction progress, the following steps are taken:

[0132] Acquire original high-resolution images of anomalous building construction. Downsample the images to reduce resolution and improve computational efficiency. Calculate the dark channel on the downsampled image and establish a minimum value table to accelerate computation. Estimate preliminary transmittance based on the dark channel results. Perform morphological edge detection on the original high-resolution image to obtain an edge binary map, and optimize the initial transmittance using the edge binary map. Measure ambient light intensity, and calculate a preliminary dehazed image using a dehazing model based on ambient light, optimized transmittance, and the original image. Upsample the preliminary dehazed image to the original resolution and process it to obtain the final dehazed image. Collect a dataset of anomalous building construction images containing different construction stages and annotate key building element data. Train a SegNet network model using the annotated data. Use the trained SegNet model to predict the semantic segmentation results of the final dehazed image to obtain semantic segmentation results of building elements. Evaluate the current construction progress information based on the presence of different building elements in the segmentation results.

[0133] In the above example, key progress information was successfully extracted from abnormal construction images, ensuring accuracy in complex and challenging environments. Preferably, the steps of calculating the dark channel on the downsampled image, establishing a minimum value table to accelerate the calculation, and estimating the preliminary transmittance of the downsampled image based on the dark channel results include the following steps:

[0134] S431. Normalize the downsampled image and map it to a set range;

[0135] S432. Slide a window in the downsampled image. In each window, traverse all pixels and find the minimum value of the red, green and blue channels. (Sliding window is a common image processing technique used to move and analyze an image in steps. In this step, find the minimum value of the three color channels (RGB) in each window.)

[0136] S433. Take the minimum value of the red, green and blue channels of each window as the dark channel value of that window (the concept of dark channel is based on the observation that in most natural scenes, at least one color channel has a very small value, which is usually related to the shadow or dark part of the object surface. By taking the minimum value of the three channels in each window, this dark part information can be found approximately).

[0137] S434. Combine the dark channel values ​​of the red, green, and blue channels to obtain the dark channel map of the downsampled image (combine the dark channel values ​​of each window to form a single dark channel map, which will be used for the estimation of atmospheric light value and transmittance in subsequent steps).

[0138] S435. Select the lowest brightness empirical value pixel from the dark channel image (it is generally believed that the darkest pixel in the dark channel image corresponds to the darkest part of the scene because the influence of atmospheric scattering or fog is minimal, so it can be used as an empirical value pixel to estimate the atmospheric light value).

[0139] S436. Find the brightness value corresponding to the empirical value pixel in the original downsampled image, take the average value as the global atmospheric light value, and calculate the preliminary transmittance of the downsampled image (by finding the pixel corresponding to the darkest pixel selected in the dark channel map in the original image, estimate the global atmospheric light value, which reflects the lighting conditions in the scene, and then use this atmospheric light value and the dark channel map to calculate the preliminary transmittance, which reflects the sharpness or fog concentration of each pixel in the image).

[0140] The formula for calculating the initial transmittance of the downsampled image is as follows:

[0141]

[0142] In the formula, ω Indicates empirical weights;

[0143] J represents the dark channel value;

[0144] A represents the global atmospheric value;

[0145] K represents the initial transmittance of the downsampled image. Preferably, the step of performing morphological edge detection on the original high-resolution abnormal building construction image to obtain an edge binary map, and then using the edge binary map to optimize the initial transmittance of the abnormal building construction image to obtain the optimized transmittance of the abnormal building construction image includes the following steps:

[0146] S441. Perform high-pass filtering on the original high-resolution abnormal building construction image to filter out uniform areas and preserve edges and details (the pass filter allows high-frequency signals (such as edges and details) to pass through while suppressing low-frequency signals (such as uniform areas), thus enhancing edge information in the image).

[0147] S442. Perform Sobel edge detection on the filtered image to detect image edges (Sobel operator is a common edge detection algorithm that detects edges by calculating the spatial gradient of image brightness. In the filtered image, the edge information is more obvious, which enables Sobel operator to effectively detect edges).

[0148] S443. Perform double thresholding on the Sobel edge detection results to obtain a binary edge map (double thresholding is an image segmentation technique that uses two thresholds to divide edge pixels into strong edges and weak edges);

[0149] S444. Perform dilation and erosion operations on the edge binary image to remove noise (dilation and erosion are morphological operations used to smooth image edges, connect broken edge segments, and remove noise points).

[0150] S445. Use the initial transmittance of the downsampled image as the initial transmittance of the abnormal building construction image.

[0151] S446. For each pixel in a non-edge region, calculate the weight based on the spatial distance between it and neighboring pixels in a preset neighborhood, and calculate the weighted average of the transmittance in that neighborhood. Set the transmittance of the pixel to the calculated weighted average transmittance to adjust the transmittance of the non-edge region (by weighting the transmittance of each non-edge pixel in its neighborhood to adjust the transmittance, the transmittance is made smoother in space, reducing the impact of local outliers).

[0152] S447. Repeat step S446 until the transmittance of all non-edge regions is smoothed to the average value within the neighborhood, and the optimized original image transmittance is obtained.

[0153] Preferably, the step of calculating a weight for each pixel in a non-edge region based on its spatial distance to neighboring pixels within a preset neighborhood, calculating a weighted average of the transmittance within that neighborhood, and setting the transmittance of the pixel to the calculated weighted average transmittance to adjust the transmittance of the non-edge region includes the following steps:

[0154] S4461. Define the neighborhood size. For each pixel in a non-edge region, traverse each pixel within its defined neighborhood.

[0155] S4462. Calculate the spatial distance between a pixel and each neighboring pixel using Euclidean distance.

[0156] S4463. Set the influence weight of each pixel, where the weight is inversely proportional to the distance;

[0157] S4464. For each neighboring pixel, multiply the transmittance by the corresponding weight, then sum all the products to get the weighted transmittance. Next, calculate the sum of all weight values, divide the weighted transmittance by the weight sum, and calculate the weighted average transmittance of the pixels in the neighborhood.

[0158] S4465. Set the transmittance of the pixel to the weighted average transmittance obtained in the previous step.

[0159] The formula for calculating the spatial distance between a pixel and each pixel in its neighborhood using Euclidean distance is as follows:

[0160]

[0161] In the formula, (x1, y1) represents the coordinates of the current pixel.

[0162] (x2, y2) represents the coordinates of a pixel within the neighborhood. Preferably, the step of measuring ambient light intensity and calculating the preliminary dehazed image using a dehazing model based on ambient light, the optimized transmittance of the abnormal building construction image, and the original high-resolution abnormal building construction image includes the following steps:

[0163] S451. In the original high-resolution abnormal building construction image, select the brightest empirical value pixel in the image.

[0164] S452. Calculate the average value of the red, green, and blue channels in the brightest empirical value pixel as the global ambient light intensity;

[0165] S453. According to the dehazing formula, traverse all pixels. For each pixel, subtract the ambient light intensity from the color value of the pixel in the red, green and blue channels to remove the influence of ambient light on the image.

[0166] S454. Add the ambient light intensity back to the red, green, and blue channels to restore the color values ​​of the pixels after dehazing, and obtain a preliminary dehazed image.

[0167] Preferably, the step of using the trained SegNet network model to predict the final dehazed image and obtain the semantic segmentation result of the building elements includes the following steps:

[0168] S491. Normalize the final dehazed image (normalization is to eliminate the influence of scale differences between different images, which usually involves scaling the pixel values ​​of the image to a fixed range, such as 0 to 1).

[0169] S492. Input the processed final dehazed image into the trained SegNet model for forward computation (forward computation means passing the input image through the network layers one by one until the last layer outputs the result).

[0170] S493. The encoder of the SegNet model extracts the features of the image, and the decoder upsamples the feature map to the original image size (the encoder gradually reduces the size of the image through convolutional and pooling layers while extracting image features, and the decoder gradually restores the image size through upsampling operations while combining the feature maps of the corresponding layers in the encoder).

[0171] S494. While upsampling the decoder, the feature map corresponding to the encoder is passed to the decoder through skip connections. Then, the output feature map is classified by softmax to obtain the probability of each pixel belonging to each category (skip connections allow the decoder to directly access the feature map in the early stages of the encoder, thus retaining more detailed information. Softmax classification transforms the feature map into a probability distribution, with each pixel corresponding to a category probability).

[0172] S495. Based on the category with the highest probability, classify each pixel to generate semantic segmentation results for building elements;

[0173] S496. Based on the segmentation results, calculate the existence ratio and distribution information of various building elements (by analyzing the segmentation results, the quantity and distribution of various building elements can be statistically determined, and the information is used to assess the construction progress).

[0174] S497. Visualize and compare the segmentation results with the original image, and check the segmentation effect.

[0175] Preferably, the step of evaluating the current construction progress information based on the presence of different building elements in the segmentation results includes the following steps:

[0176] S4101. Determine the correspondence between different building elements and building stages (link building elements (such as foundation, walls, roof, etc.) with different stages of construction (such as piling, wall construction, roofing, etc.);

[0177] S4102. Statistical analysis of the segmentation results to determine the proportion of different building elements and their area ratio (by analyzing the segmentation results, calculate the number of times each building element appears in the image and the total area it occupies).

[0178] S4103. Based on the statistical results, make a preliminary judgment on the current construction stage of the building (by comparing the proportion of different building elements and their area ratio with the construction stage, infer the approximate construction stage of the building).

[0179] S4104. Compare the current construction stage with the planned stage. If the current stage is behind the planned stage, then there is a delay in the progress.

[0180] If the current stage is ahead of the planned stage, there is an error in the schedule, and the quality needs to be rechecked;

[0181] S4105. Determine the current progress percentage based on the completion status of different building elements, and give an assessment result of the current construction progress by comprehensively considering factors such as building elements, the corresponding proportion of the workload, and the quality status (considering the complexity of building elements, the size of the workload, and the quality standards, and comprehensively assessing the current progress).

[0182] S4106. Compare the evaluation results with historical data and planned progress to determine the progress deviation.

[0183] Furthermore, assuming there is a 3x3 neighborhood and a 5x5 image, and we want to perform a neighborhood traversal on the center pixel (coordinates (2, 2)), the defined neighborhood size is 3x3, and we will consider all pixels within the 3x3 region surrounding the center pixel;

[0184] The center pixel has coordinates (2, 2) and a value of 18.

[0185] Now we need to traverse the 3x3 neighborhood around the center pixel:

[0186] Within this neighborhood, each pixel will be traversed, and subsequent steps (such as calculating Euclidean distance, weights, etc.) will be performed.

[0187] For each neighboring pixel, follow these steps:

[0188] Calculate the Euclidean distance between the current pixel and every pixel in its neighborhood. For example, for the center pixel (2, 2) and its neighboring pixel (1, 1), the Euclidean distance is sqrt((2-1)). 2 +(2-1) 2 = sqrt(2), where sqrt represents finding the square root of a number.

[0189] The influence weight of each pixel is calculated based on the Euclidean distance. The weight is inversely proportional to the distance, i.e., weight = 1 / distance.

[0190] Multiply the transmittance of each neighboring pixel by its weight. For example, if the transmittance of a neighboring pixel is 0.5 and its weight is 0.5, the product is 0.25.

[0191] Summing all the products yields the weighted sum of the transmittances. For example, if there are three neighboring pixels whose transmittance products with weights are 0.25, 0.5, and 0.25 respectively, the sum is 1.

[0192] Calculate the sum of all weight values. For example, if the weights of three neighboring pixels are 0.5, 1, and 0.5, the sum is 2.

[0193] Divide the sum of weighted transmittances by the sum of weights to obtain the weighted average transmittance of the pixels in the neighborhood. For example, if the sum of weighted transmittances is 1 and the sum of weights is 2, then the weighted average transmittance is 0.5.

[0194] Set the transmittance of the center pixel to the weighted average transmittance obtained in the above steps.

[0195] It should be explained that architectural images taken under abnormal weather conditions require preprocessing such as dehazing to improve image quality before further analysis. Dark channel prior is an effective dehazing method. Deep learning models such as SegNet are suitable for pixel-level semantic segmentation of architectural images, automatically identifying different architectural elements. A training set containing images from different construction stages can be constructed to train the SegNet model to recognize architectural elements. Engineering experience knowledge can be used to correlate architectural elements with construction progress stages to assess the current stage. Information such as the proportion of architectural elements reflects the progress completion status, and combined with quantity calculations, the progress percentage can be determined. Comparing the current progress with the planned progress and historical data can identify progress deviations. The evaluation results are visualized and reports are generated, providing feedback to project managers to optimize planning and schedule control. Compared with traditional methods, this technology achieves intelligent and automated construction progress monitoring.

[0196] SegNet is a semantic segmentation model based on convolutional neural networks (CNNs) for pixel-level image segmentation. It employs an encoder-decoder structure, where the encoder extracts image features, and the decoder maps those features back to the original image size. SegNet performs exceptionally well in many computer vision tasks, particularly image segmentation.

[0197] Dark channel dehazing is a common image dehazing algorithm based on the observation that in natural images, any region of an opaque object contains at least one darker pixel, called a dark channel pixel. Dark channel dehazing algorithms utilize this observation, removing haze from an image by estimating the dark channel pixels and the global ambient light intensity.

[0198] S5. Compare key construction progress information with actual progress information, planned progress information, and historical progress information from the previous period to comprehensively assess progress deviations.

[0199] It's important to explain that by comparing actual progress information, planned progress information, and historical progress information from the previous period, a comprehensive assessment of project schedule deviation can be achieved. Actual progress information is determined based on the actual amount of work completed and the time spent; planned progress information is determined based on the schedule outlined in the project plan; and historical progress information is determined based on the actual progress of similar projects during the same period in the past. A comprehensive assessment of schedule deviation can be achieved by calculating deviation indicators, such as schedule deviation rate, schedule deviation index, and schedule deviation degree. These indicators help project managers understand the gap between the actual and planned progress and provide a basis for developing adjustment plans.

[0200] S6. Visualize the construction progress assessment results based on construction images through 3D modeling, and generate a progress assessment report containing images, data analysis and recommendations.

[0201] It's important to explain that 3D modeling is a technology that uses computer software to create three-dimensional objects or scenes. It can generate realistic architectural models based on design drawings, data, and parameters. In construction projects, specialized modeling software can be used to create architectural models that reflect the actual situation of the project. Construction progress assessment based on construction images is the process of evaluating and analyzing the progress of a construction project. By comparing the actual progress with the planned progress, schedule deviations can be identified, and corresponding suggestions and measures can be proposed to ensure the project is completed on time. Visualizing the results of construction progress assessment based on construction images provides a more intuitive understanding of the project's actual situation. 3D modeling technology combines actual progress information with design drawings to generate a realistic architectural model. Visualizing the model allows for a clearer view of the completion status of architectural elements and the location of schedule deviations. A progress assessment report is a document generated based on the results of construction progress assessment based on construction images, providing information on the project's actual progress, data analysis, and recommendations. The report typically includes viewpoint screenshots of the architectural model, detailed data analysis of schedule deviations, and the completion status of key milestones. The report helps project managers better understand the project's progress and make corresponding decisions and plans.

[0202] In summary, by utilizing the above-mentioned technical solutions of this invention, meteorological stations and drones can collect meteorological parameter data and images in real time to monitor the weather conditions at construction sites. This allows for timely acquisition of the latest weather information, providing decision-making support for site managers. By establishing time-series models and extracting key features using meteorological parameter data and images, a comprehensive forecast of the weather for the coming week can be made. This enables advance weather planning and the implementation of corresponding measures to address potential abnormal weather. Real-time wireless communication between the meteorological station and the drone control station allows for the development of drone flight plans under normal weather conditions based on the comprehensive forecast results, improving monitoring efficiency and reducing the waste of human resources. Furthermore, this invention effectively improves the clarity and contrast of abnormal construction images through dark channel defogging technology, making details in the images more apparent and providing a better foundation for subsequent analysis and processing. Using the SegNet deep learning network, key building elements can be accurately extracted from the defogging images, providing a basis for construction image-based analysis and processing. This invention provides accurate data support for progress assessment. It is applicable not only to conventional construction images but also to images under abnormal conditions, such as fog, rain, and mist. Through morphological edge detection and transmittance optimization, it further improves image quality, making architectural elements in the images more distinct. Beyond image quality enhancement, it can also extract key construction progress information from the images, providing strong data support for construction management and monitoring. This invention can accurately identify and analyze key information in construction images, providing more precise image-based construction progress assessment. Through automated image processing and analysis, it can acquire construction progress information in real time, promptly identifying deviations and problems. Compared to traditional manual assessment, this invention is more objective and accurate, reducing human interference. The assessment method can be adjusted and optimized according to different construction projects and conditions to meet diverse needs. By comparing with historical data, it can predict future construction progress and potential problems, providing forward-looking suggestions for construction management.

[0203] Although the present invention has been disclosed above with reference to preferred embodiments, the embodiments are merely examples for illustrative purposes and are not intended to limit the present invention. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of the present invention. The scope of protection claimed by the present invention should be determined by the claims.

Claims

1. A method for assessing construction progress based on construction images, characterized in that, This construction progress assessment method based on construction images includes the following steps: S1. Set up a meteorological station at the construction site and equip it with a drone and a wireless communication module for the meteorological station; S2. The meteorological station predicts the weather conditions at the construction site for the next week, obtains meteorological parameter data and meteorological images, makes a comprehensive forecast of the future weather based on the meteorological parameter data and meteorological images, and classifies the weather conditions into normal weather and abnormal weather based on the comprehensive forecast results. S3. Develop flight plans for abnormal weather conditions based on comprehensive weather forecasts and acquire images of abnormal building construction. S4. Improve the quality of abnormal building construction images by using dark channel dehazing, and extract key information on construction progress from the enhanced abnormal building construction images using the SegNet deep network. S5. Compare key construction progress information with actual progress information, planned progress information, and historical progress information from the previous period to comprehensively assess progress deviations. S6. Visually present the construction progress assessment results based on construction images through 3D modeling, and generate a progress assessment report containing images, data analysis and recommendations; The method of improving the quality of abnormal building construction images by using dark channel dehazing and extracting key information about construction progress from the enhanced abnormal building construction images using the SegNet deep network includes the following steps: S41. Obtain the original high-resolution abnormal building construction image from the abnormal building construction image; S42. Downsample the original high-resolution abnormal building construction image to obtain a downsampled image with reduced resolution; S43. Calculate the dark channel on the downsampled image, establish a minimum value table to accelerate the calculation, and estimate the preliminary transmittance of the downsampled image based on the dark channel results. S44. Perform morphological edge detection on the original high-resolution abnormal building construction image to obtain an edge binary map, and use the edge binary map to optimize the initial transmittance of the abnormal building construction image to obtain the optimized transmittance of the abnormal building construction image. S45. Measure the ambient light intensity, and calculate the preliminary dehazing image based on the ambient light, the transmittance of the optimized abnormal building construction image, and the original high-resolution abnormal building construction image using the dehazing model. S46. Upsample the initial dehazed image to obtain a dehazed image of the same size as the original high-resolution abnormal building construction image, and process it to obtain the final dehazed image; S47. Collect a dataset of abnormal building construction images containing different building stages, and annotate key building element data; S48. Train a SegNet network model using labeled key building element data. S49. Use the trained SegNet network model to predict the final dehazed image and obtain the semantic segmentation results of the building elements. S410. Based on the presence of different building elements in the segmentation results, assess the current construction progress information.

2. The construction progress assessment method based on construction images according to claim 1, characterized in that, The meteorological station forecasts the weather conditions at the construction site for the next week, acquires meteorological parameter data and meteorological images, makes a comprehensive forecast of the future weather based on the meteorological parameter data and meteorological images, and classifies the weather conditions into normal weather and abnormal weather based on the comprehensive forecast results, including the following steps: S21. The meteorological station collects meteorological parameter data, which includes at least temperature data, humidity data, air pressure data, wind speed data, and wind direction data. S22. Meteorological stations collect meteorological images, which include at least cloud images, radar images, and infrared images; S23. Extract key features from meteorological parameter data and meteorological images; S24. Establish a time series model, input meteorological parameter data and key features of meteorological images, and make a comprehensive forecast of future weather to obtain the weather conditions for the next week. S25. Based on the predicted weather conditions, set thresholds to distinguish between normal weather and abnormal weather; S26. If the weather conditions are within the set threshold range, they are determined to be normal weather. If the weather exceeds the set threshold, it is considered abnormal weather.

3. The construction progress assessment method based on construction images according to claim 2, characterized in that, The meteorological station forecasts the weather conditions at the construction site for the next week, acquires meteorological parameter data and meteorological images, makes a comprehensive forecast of the future weather based on the meteorological parameter data and meteorological images, and classifies the weather conditions into normal weather and abnormal weather based on the comprehensive forecast results. This also includes the following steps: If the overall forecast indicates normal weather, then a flight plan for normal weather conditions will be formulated. Acquire construction images of the target building and extract key information about the construction progress.

4. The construction progress assessment method based on construction images according to claim 1, characterized in that, The process of calculating the dark channel on the downsampled image, establishing a minimum value table to accelerate the calculation, and estimating the preliminary transmittance of the downsampled image based on the dark channel results includes the following steps: S431. Normalize the downsampled image and map it to a set range; S432. Slide a window in the downsampled image, and in each window, traverse all pixels to find the minimum value of the red channel, green channel, and blue channel; S433. Take the minimum value of the red, green and blue channels of each window as the dark channel value of that window; S434. Combine the dark channel values ​​of the red, green, and blue channels to obtain the dark channel map of the downsampled image; S435. Select the pixel with the lowest empirical value from the dark channel map; S436. Find the brightness value corresponding to the empirical value pixel in the original downsampled image, take the average value as the global atmospheric light value, and calculate the preliminary transmittance of the downsampled image. The formula for calculating the initial transmittance of the downsampled image is as follows: In the formula, ω Indicates empirical weights; J represents the dark channel value; A represents the global atmospheric value; K represents the initial transmittance of the downsampled image.

5. The construction progress assessment method based on construction images according to claim 4, characterized in that, The process of performing morphological edge detection on the original high-resolution abnormal building construction image to obtain an edge binary map, and then using the edge binary map to optimize the initial transmittance of the abnormal building construction image to obtain the optimized transmittance of the abnormal building construction image includes the following steps: S441. Perform high-pass filtering on the original high-resolution abnormal building construction image to filter out uniform areas and retain edges and details; S442. Perform Sobel edge detection on the filtered image to detect image edges; S443. Perform double thresholding on the Sobel edge detection results to obtain a binary edge map; S444: Perform dilation and erosion operations on the edge binary image to remove noise; S445. Use the initial transmittance of the downsampled image as the initial transmittance of the abnormal building construction image. S446. For each pixel in a non-edge region, calculate the weight based on the spatial distance between it and neighboring pixels in a preset neighborhood, and calculate the weighted average of the transmittance in that neighborhood. Set the transmittance of the pixel to the calculated weighted average transmittance to adjust the transmittance of the non-edge region. S447. Repeat step S446 until the transmittance of all non-edge regions is smoothed to the average value within the neighborhood, and the optimized original image transmittance is obtained.

6. The construction progress assessment method based on construction images according to claim 5, characterized in that, The steps of adjusting the transmittance of non-edge regions by calculating the weight of each pixel in a non-edge region based on its spatial distance to neighboring pixels within a preset neighborhood, calculating the weighted average transmittance within that neighborhood, and setting the transmittance of the pixel to the calculated weighted average transmittance include the following: S4461. Define the neighborhood size. For each pixel in a non-edge region, traverse each pixel within its defined neighborhood. S4462. Calculate the spatial distance between a pixel and each neighboring pixel using Euclidean distance. S4463. Set the influence weight of each pixel, where the weight is inversely proportional to the distance; S4464. For each neighboring pixel, multiply the transmittance by the corresponding weight, then sum all the products to get the weighted transmittance. Next, calculate the sum of all weight values, divide the weighted transmittance by the weight sum, and calculate the weighted average transmittance of the pixels in the neighborhood. S4465. Set the transmittance of the pixel to the weighted average transmittance obtained in the previous step. The formula for calculating the spatial distance between a pixel and each pixel in its neighborhood using Euclidean distance is as follows: In the formula, (x1, y1) represents the coordinates of the current pixel. (x2, y2) represents the coordinates of a pixel in the neighborhood.

7. The construction progress assessment method based on construction images according to claim 6, characterized in that, The process of measuring ambient light intensity and calculating a preliminary dehazed image using a dehazing model based on ambient light, the optimized transmittance of the abnormal building construction image, and the original high-resolution abnormal building construction image includes the following steps: S451. In the original high-resolution abnormal building construction image, select the brightest empirical value pixel in the image. S452. Calculate the average value of the red, green, and blue channels in the brightest empirical value pixel as the global ambient light intensity; S453. According to the dehazing formula, traverse all pixels. For each pixel, subtract the ambient light intensity from the color value of the pixel in the red, green and blue channels to remove the influence of ambient light on the image. S454. Add the ambient light intensity back to the red, green, and blue channels to restore the color values ​​of the pixels after dehazing, and obtain a preliminary dehazed image.

8. The construction progress assessment method based on construction images according to claim 7, characterized in that, The process of using a trained SegNet network model to predict the final dehazed image and obtain the semantic segmentation results of building elements includes the following steps: S491. Normalize the final dehazed image after dehazing. S492. Input the processed final dehazed image into the trained SegNet model for forward computation; S493. The encoder of the SegNet model extracts the features of the image, and the decoder upsamples the feature map to the original image size. S494. While upsampling the decoder, the feature map corresponding to the encoder is passed to the decoder through a skip connection. Then, the output feature map is classified by softmax to obtain the probability of each pixel belonging to each category. S495. Based on the category with the highest probability, classify each pixel to generate semantic segmentation results for building elements; S496. Based on the segmentation results, calculate the existence ratio and distribution information of various architectural elements; S497. Visualize and compare the segmentation results with the original image, and check the segmentation effect.

9. A construction progress assessment method based on construction images according to claim 8, characterized in that, The process of assessing the current construction progress information based on the presence of different building elements in the segmentation results includes the following steps: S4101. Determine the correspondence between different architectural elements and architectural stages; S4102. Information on the proportion of different building elements and their area ratio in the statistical segmentation results; S4103. Based on the statistical results, make a preliminary judgment on the current construction stage of the building; S4104. Compare the current construction stage with the planned stage. If the current stage is behind the planned stage, then there is a delay in the progress. If the current stage is ahead of the planned stage, there is an error in the schedule, and the quality needs to be rechecked; S4105. Determine the current progress percentage based on the completion status of different building elements, and give an assessment result of the current construction progress by comprehensively considering factors such as building elements, corresponding project volume proportion, and quality status. S4106. Compare the evaluation results with historical data and planned progress to determine the progress deviation.

Images