A rapid detection method and device for the coating state of milled asphalt pavement material based on multispectral imaging

Abstract

This invention discloses a rapid detection method and device for the coating state of milled asphalt pavement aggregate based on multispectral imaging, belonging to the field of road material testing technology. The device includes a sample carrying and transport module, a multispectral imaging module, and a data processing and control module. This invention acquires multispectral images of milled aggregate in the visible-near-infrared band under multi-light source conditions, and uses image classification algorithms to automatically and accurately identify and distinguish asphalt-coated areas from exposed aggregate areas, thereby calculating quantitative indicators such as asphalt coating rate and coating uniformity index. This invention effectively solves the problems of strong subjectivity and poor accuracy in traditional manual inspection, achieving an objective, rapid, and quantitative evaluation of the quality of milled aggregate.

Description

Technical Field

[0001] This invention relates to the field of road material testing technology, and in particular to a rapid and quantitative testing device for the coating state of reclaimed asphalt pavement (RAP) based on multispectral imaging technology. Background Technology

[0002] In asphalt pavement recycling technology, the large-scale and high-value utilization of milled aggregate is key to the industry's sustainable development. The performance of RAP (Rinse-Asphalt Pavement) largely depends on the coating state of its surface aged asphalt, including coating rate, uniformity, and degree of aging. RAP with poor coating is difficult to effectively integrate with new asphalt during the recycling process, easily leading to problems such as insufficient strength, reduced water damage resistance, and poor durability in the recycled mixture.

[0003] Currently, the evaluation of RAP coating status largely relies on manual visual observation and experience-based judgment, which is highly subjective, has poor repeatability, and lacks precise quantitative indicators. Although some studies have attempted to use digital image processing (DIP) technology for analysis, most of these studies are conducted under a single white light source, making them highly susceptible to interference from the color, texture, dust, and shadows on the RAP surface. This makes it impossible to reliably distinguish between asphalt film and exposed aggregate, resulting in limited detection accuracy and reliability.

[0004] In recent years, hyperspectral and multispectral imaging technologies have proven to have significant advantages in distinguishing asphalt from aggregates due to their ability to capture the differences in spectral characteristics of materials in specific wavelength bands. Related studies have shown that asphalt and different types of aggregates have unique spectral responses in the visible-near infrared (VIS-NIR, 400-1000nm) and short-wave infrared (SWIR, 900-1700nm) bands. Combined with image classification algorithms (such as ISODATA and parallelelepiped), pixel-level accurate classification can be achieved, thereby objectively quantifying the asphalt coating rate and effectively overcoming the subjectivity of manual interpretation.

[0005] However, existing technologies are mostly limited to laboratory analysis and have not yet formed a rapid testing device that integrates automated sample delivery, stable optical environment control, multispectral imaging, and online data processing. Therefore, developing a dedicated device that can provide objective, rapid, and accurate quantitative evaluation of RAP quality is crucial for ensuring the quality of recycled asphalt mixtures and promoting their standardized application. Summary of the Invention

[0006] Purpose of the invention

[0007] The purpose of this invention is to overcome the shortcomings of existing technologies, such as strong subjectivity, low efficiency, and insufficient accuracy, and to provide an integrated, automated, and quantitative rapid detection device for the coating status of milled asphalt pavement materials.

[0008] Technical solution

[0009] To achieve the above objectives, this invention provides a rapid detection method and apparatus for the coating state of milled asphalt pavement material based on multispectral imaging, which mainly includes:

[0010] 1. Sample Carrying and Transport Module: Used to carry and smoothly transport the milled material sample to be tested through the testing area. This module includes:

[0011] (1) Feed inlet: used to feed RAP samples onto the conveyor belt.

[0012] (2) Vibrating cloth spreader: used to spread RAP samples evenly and avoid stacking.

[0013] (3) Uniform speed conveyor belt: used to carry and transport samples distributed in a single layer at a constant speed to ensure imaging consistency.

[0014] (4) Collection box: Collect the RAP samples after the test is completed.

[0015] 2. Multispectral imaging module: Directly facing the detection area. This module includes:

[0016] (1) Multispectral camera: Preferably a camera with visible light and near-infrared imaging capabilities, with a spectral range covering 400-1700nm, used to acquire high spectral resolution images of milled material samples.

[0017] (2) Controllable multi-angle illumination unit: arranged around the camera, including multiple wavelength LED light sources (such as white light, 850nm, 940nm, etc.), which can be switched according to a preset sequence to provide uniform, multi-angle illumination to stimulate the differential spectral response of the sample surface at different wavelengths.

[0018] 3. Data processing and control module: Electrically connected to the multispectral imaging module and the sample carrying and transport module, this module is configured to perform the following functions:

[0019] (1) Control unit: used to coordinate the wavelength switching of the illumination unit, the start and stop and speed of the conveyor belt, and the shooting trigger of the multispectral camera to realize the automation of the detection process.

[0020] (2) Image Processing Unit: Receives multispectral image data and performs preprocessing (including radiometric calibration, geometric correction, and denoising). Subsequently, an image classification algorithm is used to perform pixel-level classification on the preprocessed image, accurately distinguishing between "asphalt-coated areas" and "exposed aggregate areas." The classification algorithm includes, but is not limited to, supervised classification algorithms (such as parallelelepiped classifiers and support vector machines) or unsupervised classification algorithms (such as ISODATA).

[0021] (3) State Analysis Unit: Based on the image classification results, calculate one or more quantitative indicators characterizing the asphalt coating state, including: ① Asphalt Coating Rate: Calculated based on the ratio of the total number of pixels classified as "asphalt-coated areas" to the total number of pixels in the entire detection area. ② Coating Uniformity Index: Calculated based on the spatial distribution statistical characteristics of the pixels in the "asphalt-coated areas" (such as the coefficient of variation and gray-level co-occurrence matrix texture features), used to evaluate the uniformity of coating.

[0022] Furthermore, the spectral range of the multispectral imaging module covers the VIS-NIR and SWIR bands to fully utilize the differences in spectral absorption characteristics between asphalt and aggregate in this region.

[0023] Furthermore, the radiometric calibration uses a standard white board (such as Spectralon) for relative reflectivity calibration to eliminate the effects of instrument response and uneven illumination.

[0024] Beneficial effects

[0025] Compared with the prior art, the present invention has the following significant advantages:

[0026] 1. High detection accuracy and strong anti-interference ability: By utilizing the spectral recognition ability of multispectral imaging technology in specific bands (especially VIS-NIR and SWIR), it can effectively capture the spectral feature differences between asphalt and aggregates, thereby accurately distinguishing the two and overcoming the interference of factors such as color, shadow, and surface dust, achieving a truly objective quantitative evaluation.

[0027] 2. Fully automated process with significantly improved efficiency: From automatic sample preparation and delivery to synchronous acquisition of multispectral images, and then to automatic data processing and result output, the entire testing process can be completed within tens of seconds without manual intervention, greatly improving testing efficiency and making it suitable for the rapid needs of industrial production and quality control.

[0028] 3. Comprehensive evaluation dimensions and more instructive results: It not only provides the overall asphalt coating rate, but also calculates the spatial distribution index that characterizes the uniformity of coating, providing richer and more in-depth quantitative basis for the comprehensive evaluation of RAP quality.

[0029] 4. Wide applicability and good robustness: By adopting different image classification algorithms (such as ISODATA for exploration without prior knowledge and parallelelepiped for training with specific samples), this device can adapt to different types and colors of aggregates (such as limestone, basalt, etc.), and has good versatility and robustness.

[0030] 5. Excellent cost-effectiveness and easy to promote: Compared with the more complex and expensive hyperspectral imaging system, the multispectral imaging combined with advanced digital image processing technology of this invention can effectively control costs while ensuring high accuracy, making it easier to promote and apply in production and quality inspection units. Attached Figure Description

[0031] Figure 1 This is a three-dimensional structural diagram of the invention from a first perspective.

[0032] Figure 2 This is a three-dimensional structural diagram of the invention from a second perspective.

[0033] Figure 3 This is a three-dimensional structural diagram of the vibrating cloth feeder in this invention.

[0034] Figure 4 This is an enlarged view of the controllable multi-angle lighting unit in this invention.

[0035] Figure 5 This is a flowchart of the present invention.

[0036] Among them, 1-feed inlet; 2-conveyor belt; 3-vibrating cloth distributor; 31-conveyor shaft; 32-spring ball; 4-collection frame; 5-multispectral camera; 6-controllable multi-angle lighting unit; 61-white light; 62-near-infrared LED; 63-short-wave infrared LED; 7-industrial computer; 8-motor; 9-belt; 10-hub. Detailed Implementation

[0037] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0038] like Figure 1 and Figure 2 As shown, the device of the present invention mainly consists of three parts: a sample carrying and transport module, a multispectral imaging module, and a data processing and control module.

[0039] RAP samples are fed into a conveyor belt through the inlet. A vibrating distributor evenly disperses the RAP samples onto the constant-speed conveyor belt, forming a single-layer particle flow. The multispectral imaging module is fixed to a top support, with its optical axis perpendicular to the conveyor belt plane. A multispectral camera (e.g., a snapshot camera with VIS-NIR sensing capability) is responsible for acquiring images. A controllable multi-angle illumination unit surrounding the camera consists of a ring of LEDs with multiple wavelengths, which are sequentially triggered by a control unit to achieve multi-band image acquisition.

[0040] The core of the data processing and control module is an industrial computer. Its workflow is as follows:

[0041] 1. System initialization: RAP samples enter the detection area after passing through a vibrating fabric.

[0042] 2. Trigger the sensor to send a signal to the control unit.

[0043] 3. The control unit instructs the illumination unit to switch wavelengths according to a preset sequence and simultaneously triggers the multispectral camera to acquire a series of multispectral images.

[0044] 4. The image processing unit preprocesses the acquired image (including radiometric calibration based on whiteboard and dark current), and then uses a classification algorithm (such as parallelepiped supervised classification) to classify the image pixels and identify "asphalt coating" and "exposed aggregate".

[0045] 5. The state analysis unit counts the number of pixels of various types and calculates the bitumen coating rate (BIT). At the same time, by analyzing the spatial distribution characteristics of the "bitumen coating" pixels, the coating uniformity index is calculated.

[0046] 6. The final results are displayed on the interface and a test report is generated.

Claims

1. A rapid detection method and device for the coating state of milled asphalt pavement material based on multispectral imaging, characterized in that, include: (1) Sample carrying and conveying module, used to make the milling material sample to be tested pass through the detection area smoothly in a single layer; (2) Multispectral imaging module, used to acquire multispectral image data of asphalt pavement milling material samples passing through the detection area, the spectral range of which covers the visible light to near-infrared band; (3) The data processing and control module is electrically connected to the multispectral imaging module and the sample carrying and transporting module, and is configured to perform the following operations: control the collaborative work of the sample carrying and transporting module and the multispectral imaging module; process the multispectral image data, and distinguish between the asphalt-coated area and the exposed aggregate area through an image classification algorithm; and calculate a quantitative index characterizing the asphalt coating state based on the distinction result.

2. The apparatus according to claim 1, characterized in that, The sample carrying and conveying module includes ① a feed inlet, ② a uniform speed conveyor belt, ③ a vibrating spreader and ④ a collection frame. The ③ vibrating spreader is used to evenly spread the milled material sample on the ② uniform speed conveyor belt.

3. The apparatus according to claim 1, characterized in that, The multispectral imaging module includes a ⑤ multispectral camera and a ⑥ controllable multi-angle illumination unit. The ⑥ controllable multi-angle illumination unit can provide three different wavelengths of light source and can switch them according to a preset sequence.

4. The apparatus according to claim 3, characterized in that, The spectral range of the multispectral imaging module covers 400nm to 1700nm.

5. The apparatus according to claim 3, characterized in that, The controllable multi-angle lighting unit ⑥ includes white LEDs and two types of near-infrared LEDs.

6. The apparatus according to claim 1, characterized in that, The image classification algorithm is a supervised classification algorithm or an unsupervised classification algorithm; preferably, the supervised classification algorithm is a parallelepiped classifier or a support vector machine, and the unsupervised classification algorithm is the ISODATA algorithm.

7. The apparatus according to claim 1, characterized in that, Each pixel in the multispectral image contains a continuous spectral curve from VIS to SWIR. The data processing and control module analyzes the spectral curve of each pixel using the image classification algorithm and assigns a category label based on its similarity to the spectral characteristics of "typical asphalt" or "typical aggregate," thereby distinguishing asphalt-coated areas from exposed aggregate areas. The asphalt coating rate is calculated based on the distinction results using the following formula: Asphalt coating rate = (number of pixels classified as asphalt-coated areas / total number of pixels in the detection area) × 100%.

8. The apparatus according to claim 7, characterized in that, The quantitative index characterizing the asphalt coating state also includes the coating uniformity index, which is calculated based on the spatial distribution statistical characteristics of the pixels in the asphalt coating area.

9. The apparatus according to claim 1, characterized in that, Before performing image processing, the data processing and control module first performs radiometric calibration preprocessing on the multispectral image data. The radiometric calibration uses a standard white board to calibrate the relative reflectance.

Images