A high-precision airborne remote sensing method for monitoring sea drift garbage based on polarization filtering

Abstract

The present application discloses a high-precision airborne remote sensing sea drift garbage monitoring method based on polarization filtering. The steps include: collecting a remote sensing image set of the sea drift garbage area; preprocessing the remote sensing image set to obtain a processed image set; constructing a spatio-temporal information sequence based on the processed image set; performing garbage recognition and garbage annotation on the processed image set; obtaining the information of the garbage according to the spatio-temporal information sequence and sending the information to the cleaning personnel to complete garbage cleaning. By using polarization imaging, the present application filters out the reflected light of the solar flare caused by the undulation of the sea surface waves, reduces the interference light parameters of the solar flare reflection in the remote sensing image, and improves the monitoring accuracy. At the same time, by constructing the spatio-temporal information sequence of the sea surface, the position of the garbage can be quickly and rapidly locked and marked through the spatio-temporal information sequence, greatly improving the monitoring efficiency and reducing the cleaning cost.

Description

Technical Field

[0001] The present application relates to the technical field of aerial remote sensing garbage inspection, and in particular to a high-precision aerial remote sensing marine debris monitoring method based on polarization filtering. Background Art

[0002] Managing marine debris is a global marine ecological and environmental challenge. Trash from rivers entering the sea and waste generated by human activities near the coast are carried and accumulated by currents and waves, causing serious marine environmental pollution. If marine debris is not promptly managed, it can become stranded by tides, causing severe coastal pollution and damaging the coastal ecosystem. Furthermore, if not promptly cleaned up, the microplastics formed by the gradual decomposition of marine debris can enter marine life and the human body through the food chain, posing a threat to both marine life and human health.

[0003] Timely removal of marine debris is crucial. However, given the vast size of offshore waters and the high dispersion of debris, the effectiveness of monitoring its distribution directly impacts its removal. Traditional methods for monitoring marine debris rely on regular boat patrols and observations by coastal environmental protection personnel. These methods are subject to subjectivity, are time-consuming, have limited coverage, are slow to inspect, have poor horizontal visibility, make it difficult to locate debris, and leave traces after cleaning. Furthermore, high labor costs and the duplication of personnel can lead to a lack of oversight, making them unable to meet the needs of efficient environmental governance and ecological protection in vast ocean areas.

[0004] The method of inspecting marine debris based on aerial remote sensing has attracted widespread attention. Typically, aerial vehicles carrying aerial photography equipment conduct remote sensing imaging inspections of the sea area, and the distribution of marine debris is determined by identifying typical debris targets in remote sensing images. However, the reflected sun spots caused by sunlight shining on the sea surface during the day are affected by the wind and waves at sea and sunlight, resulting in irregular shapes and intensities. This greatly complicates the identification of typical debris in the image and greatly increases the misidentification rate. The method of filtering out reflected sun spots through complex image processing algorithms requires extremely strong computing power and places high demands on hardware equipment. Accurate target identification takes a long time, and image recognition takes a long time after each aerial shot, which lacks timeliness. The positioning of marine debris requires high real-time performance, which is not conducive to its promotion and application. Summary of the Invention

[0005] This application uses polarization imaging to filter out the sun spot reflection light caused by the undulations of the sea surface, reduce the sun spot reflection interference light parameters in the remote sensing image, and improve monitoring accuracy.

[0006] To achieve the above objectives, the present application provides a high-precision aerial remote sensing method for monitoring marine debris based on polarization filtering, comprising the following steps:

[0007] Collect remote sensing images of areas with floating garbage at sea;

[0008] Preprocess the remote sensing image set to obtain a processed image set;

[0009] Construct a spatio-temporal information sequence based on the processed image set;

[0010] Perform garbage recognition and garbage annotation on the processed image set;

[0011] Obtain the information of the garbage according to the spatio-temporal information sequence, and send the information to the cleaning staff to complete garbage cleaning.

[0012] Preferably, the method for obtaining the processed image set includes: using the principle of polarization imaging to filter the reflected light of the sunspots caused by the undulation of the sea surface waves, and reducing the interference light parameters of the sunspot reflections in the remote sensing image.

[0013] Preferably, the method for constructing the spatio-temporal information sequence includes: obtaining the image information continuously distributed in space in the time series based on the longitude and latitude information and the time series information in the remote sensing image set; performing fusion according to the overlapping positions of adjacent images to obtain the spatio-temporal information sequence of the aerial images of the sea area.

[0014] Preferably, the method for performing the garbage recognition includes:

[0015] Standardize the processed image set to obtain a standardized image set;

[0016] Perform information matching on the standardized image set to obtain the information that conforms to the typical sea drift garbage in the image, and complete the garbage recognition.

[0017] Preferably, the method for obtaining the standardized image set includes: converting the RGB photo into an HSV photo by using the RGB to HSV conversion method, and performing standardization on the image information by using the gamma correction method to obtain the standard image set.

[0018] Preferably, the method for obtaining the information of the typical sea drift garbage includes:

[0019] Set the HSV threshold of the typical sea drift garbage;

[0020] Compare the HSV threshold of the typical sea drift garbage with the HSV threshold of the sea surface to obtain the characteristic edge structure of the typical sea drift garbage;

[0021] Perform contour reconstruction based on the characteristic edge structure to obtain the information of the typical sea drift garbage.

[0022] Preferably, the method for obtaining the characteristic edge structure includes: according to the threshold difference between the water surface of the monitored sea area and the HSV threshold of the typical marine floating garbage, selecting the part of the image that conforms to the HSV threshold of the typical marine floating garbage in the ocean background map after filtering and standardization, and using the Harr feature pixel sub-module difference method to obtain the characteristic edge structure.

[0023] Preferably, the information of the typical marine floating garbage includes: size information, quantity information, and positioning longitude and latitude information.

[0024] Compared with the prior art, the beneficial effects of the present application are as follows:

[0025] By using polarization imaging, the present application filters out the reflected light of the sunspots caused by the undulation of the sea surface waves, reduces the interference light parameters of the sunspot reflections in the remote sensing images, and improves the monitoring accuracy. At the same time, a sea surface spatio-temporal information sequence is constructed, and the garbage position can be quickly locked and marked through the spatio-temporal information sequence, greatly improving the monitoring efficiency and reducing the cleaning cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the present application, the following briefly introduces the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0027] Figure 1 It is a schematic flowchart of the method in the embodiment of the present application;

[0028] Figure 2 It is a schematic diagram of the light path polarization of the polarization filter device in the embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0029] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0030] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0031] As Figure 1 shown, it is a schematic flowchart of the method in the embodiment of the present application, and the steps include:

[0032] S1. Collect a set of remote sensing images of the area with floating garbage at sea.

[0033] First, collect a set of remote sensing images of the area with floating garbage at sea. In this embodiment, an aircraft equipped with remote sensing equipment selects appropriate weather conditions according to meteorological and tidal forecasts for periodic collection to obtain a set of remote sensing images. In this implementation, according to meteorological and sea conditions, select weather with normal wind levels and sunlight for aerial photography of floating garbage in the sea area. The aircraft carries remote sensing imaging equipment to cruise and shoot at a specified altitude, and the shooting altitude meets the resolution conditions of the remote sensing imaging equipment.

[0034] Generally, the performance of an aerial photography remote sensing imaging device is determined by the focal length f' and the field of view angle 2ω. The focal length determines the size of the image, and the field of view determines the aerial photography imaging range. When photographing floating objects at sea, the image size y′ = -f′tanω. When the size of the photosensitive element of the remote sensing camera is fixed, the shorter the focal length of the selected imaging lens group, the larger the field of view angle of the shot. However, when applied to aerial photography, the distance between the remote sensing equipment and the sea surface is usually much larger than the focal length. Therefore, for a long-focal-length remote sensing device, the focal length is longer, and the corresponding imaging field of view angle is smaller. To achieve aerial photography remote sensing of a large sea area, for the width d of the photosensitive element of the airborne remote sensing camera, the camera focal length f', the distance D between the remote sensing aircraft and the water surface, and the length range L of the aerial photography image, then

[0035] Considering the pixel resolution of the size of floating garbage at sea on the imaging surface, usually, for floating garbage with a side length of less than 1 meter to be accurately identified on the pixel surface, it needs to occupy at least 3 pixel point lengths in a single direction. Then the relationship between the distance of the remote sensing aircraft and the water surface is

[0036] S2. Preprocess the set of remote sensing images to obtain a set of processed images.

[0037] After that, preprocess the set of remote sensing images to obtain a set of processed images. In this embodiment, use a polarization imaging system to assist the airborne remote sensing camera to filter out the reflected light of the sunspots caused by the undulation of the sea surface waves, and reduce the interference light parameters of the sunspot reflections in the remote sensing images.

[0038] When an aircraft carries a remote sensing aerial photography device to conduct aerial photography within a specified altitude range, the resolution of the target object in the remote sensing image can be guaranteed. The main factor affecting the accuracy of sea drift garbage recognition is the irregular-shaped sunlight spots caused by the undulation of sea waves. Since the waves are affected by wind and tides, the intensity of sunlight changes periodically during the day, and the brightness, size, and shape of the sunlight spots reflected on the water surface are all in a dynamic change state. Moreover, typical sea drift garbage floats and fluctuates in the sea waves, and there is also a situation where it is partially submerged underwater and its shape changes dynamically. Therefore, the accuracy of extracting typical sea drift garbage through the target recognition algorithm in the image is severely affected. In this embodiment, at the front end of the aerial photography remote sensing imaging carried by the aircraft, a device for removing the interference factor of sunlight spots based on polarization filtering is added to preprocess the collected images, reducing the interference factors from the image end, and thus improving the accurate recognition rate and image processing efficiency of the entire system.

[0039] As Figure 2 shown, the polarization filter device adopted in this embodiment is composed of a combination of polarization lenses matching the aperture of the imaging lens, including a linear polarizer and a quarter-wave plate. The optical signal entering the polarization filter device is mainly composed of two parts: the external ambient light and the light reflected by the sea surface.

[0040] Among them, the external ambient light is mainly the scattered sunlight. Sunlight is approximately completely polarized light, and the light vectors are evenly distributed in all directions and have the same amplitude. However, the reflectivity of the sea surface to the light vectors in different polarization directions of sunlight is different. After being reflected by the sea surface, it can be linearly reflected perpendicular to the incident plane. That is, in this specific direction, there will be a strong distribution of vibrations of polarized light, and the light vibration perpendicular to the incident plane in the reflected light is more than the parallel vibration, forming partially polarized light with a very strong polarization degree in a single direction.

[0041] When the optical signal enters the polarization filter device, it first passes through a linear polarizer, which converts the external ambient light from a completely polarized state or a partially polarized state into linearly polarized light. When the light passes through the first linear polarizer of the polarization filter device, by adjusting the angle of the linear polarizer in advance to make a difference in the polarization angle from the sea surface reflected light, most of the sunlight spots reflected on the water surface can be filtered out.

[0042] Next is a quarter-wave plate, which converts the linearly polarized light into circularly polarized light and enters the photosensitive detector of the imaging device.

[0043] The phase delay generated when the linearly polarized light passes through the quarter-wave plate can be expressed as

[0044] δ=(2m + 1)π / 2

[0045] When light passes through a quarter-wave plate, a phase delay of an odd multiple of π / 2 is generated. By controlling the fast and slow axes of the quarter-wave plate to be at ±45° with the direction of the optical vector of the incident linearly polarized light, the incident linearly polarized light can be converted into circularly polarized light, so as to be better received by the photodetector of the imaging device.

[0046] After polarization filtering, the solar light spots reflected by the water surface are basically filtered out in the aerial remote sensing image. By using the color, contour, and brightness characteristics of typical floating garbage at sea and the contrast with the seawater background, the rapid positioning and verification of typical floating garbage at sea can be achieved.

[0047] S3. Construct a spatio-temporal information sequence based on the processed picture set.

[0048] Obtain the image information continuously distributed in space in the time series according to the latitude and longitude information and the time series information in the processed picture set; perform fusion according to the overlapping positions of adjacent images to obtain the spatio-temporal image information sequence of the sea area aerial photography.

[0049] S4. Perform garbage recognition and garbage annotation on the processed picture set.

[0050] According to the color, contour, and floating characteristics of typical floating garbage at sea, use the target fast recognition algorithm to perform recognition on each image in the spatio-temporal image information sequence of the sea area. The images after interference filtering have higher accuracy and accuracy in the target fast recognition process.

[0051] In this embodiment, the picture recognition process includes: converting the RGB photo into an HSV photo by using the RGB to HSV conversion method, standardizing the image information by using the gamma correction method, matching the standardized image information, and obtaining the part of the image that meets the HSV threshold of typical sea-floating garbage; then, setting the HSV threshold of typical garbage, and according to the HSV threshold characteristics of the water surface in the monitoring sea area and the color threshold difference of typical sea-floating garbage, it is easy to select the part of the image that meets the HSV threshold of typical sea-floating garbage in the ocean background map after filtering and standardization; finally, using the Harr feature pixel sub-module difference method to obtain the fast and simple edge structure of typical sea-floating garbage features.

[0052] S5. Obtain the information of the garbage according to the spatio-temporal information sequence and send the information to the cleaning personnel to complete the garbage cleaning.

[0053] Based on the edge structure of typical sea-floating garbage features, contour reconstruction is performed. The method includes: obtaining the size category and positioning latitude and longitude of the typical sea-floating garbage for the reconstructed contour according to the pixel position and quantity information in the aerial photo and combining with the latitude and longitude information.

[0054] Subsequently, based on the typical sizes, quantities, and positioning latitudes and longitudes of the reconstructed outlines of marine debris, size information, quantity information, and positioning latitude and longitude information of the marine debris distribution will be obtained, uploaded to the control system in the form of a report, and sent to on-site cleaning personnel to achieve efficient cleaning.

[0055] The embodiments described above are only descriptions of the preferred modes of this application and do not limit the scope of this application. Without departing from the design spirit of this application, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of this application should fall within the protection scope determined by the claims of this application.

Claims

1. A high-precision airborne remote sensing method for monitoring sea drift garbage based on polarization filtering, characterized in that the steps Including: Collecting a remote sensing image set of the area of floating garbage at sea; Preprocessing the remote sensing image set to obtain a processed image set. The preprocessing includes: setting a polarization filter device composed of a linear polarizer and a quarter-wave plate at the front end of the remote sensing imaging device, and by adjusting the difference between the angle of the linear polarizer and the polarization direction of the sea surface reflected light, filtering out the specular reflection light of the sunspots caused by the undulation of the waves, and reducing the interference light parameters; Constructing a spatio-temporal information sequence based on the processed image set; Performing garbage recognition and garbage annotation on the processed image set; Obtaining the information of the garbage according to the spatio-temporal information sequence, and sending the information to the cleaning personnel to complete garbage cleaning.

2. The high-precision airborne remote sensing sea drift garbage monitoring method based on polarization filtering according to claim 1, wherein The method for obtaining the processed image set includes: using the polarization imaging principle to filter out the specular reflection light of the sunspots caused by the undulation of the sea surface, and reducing the interference light parameters of the specular reflection in the remote sensing image.

3. The high-precision airborne remote sensing sea drift garbage monitoring method based on polarization filtering according to claim 1, wherein The method for constructing the spatio-temporal information sequence includes: obtaining the image information continuously distributed in space in the time series based on the longitude and latitude information and the time series information in the remote sensing image set; performing fusion according to the overlapping positions of adjacent images to obtain the spatio-temporal information sequence of the aerial images of the sea area.

4. The high-precision airborne remote sensing sea drift garbage monitoring method based on polarization filtering according to claim 1, wherein, The method for performing the garbage recognition includes: Normalizing the processed image set to obtain a normalized image set; Performing information matching on the normalized image set to obtain the information of the typical sea drift garbage in the image, and completing the garbage recognition.

5. The high-precision airborne remote sensing sea drift garbage monitoring method based on polarization filtering according to claim 4, characterized in that, The method for obtaining the normalized image set includes: converting the RGB photo into an HSV photo by using the RGB to HSV conversion method, and performing normalization on the image information by using the gamma correction method to obtain the normalized image set.

6. The high-precision airborne remote sensing sea drift garbage monitoring method based on polarization filtering according to claim 5, wherein, The method for obtaining the information of the typical sea drift garbage includes: Setting the HSV threshold of the typical sea drift garbage; Comparing the HSV threshold of the typical sea drift garbage with the HSV threshold of the sea surface to obtain the characteristic edge structure of the typical sea drift garbage; Performing contour reconstruction based on the characteristic edge structure to obtain the information of the typical sea drift garbage.

7. The high-precision airborne remote sensing sea drift garbage monitoring method based on polarization filtering according to claim 6, characterized in that, The method for obtaining the characteristic edge structure includes: selecting the part of the image that conforms to the HSV threshold of the typical sea drift garbage in the filtered and normalized ocean background map according to the threshold difference between the water surface of the monitored sea area and the HSV threshold of the typical sea drift garbage, and using the Harr feature pixel sub-module difference method to obtain the characteristic edge structure.

8. The high-precision airborne remote sensing sea drift garbage monitoring method based on polarization filtering according to claim 7, characterized in that, The information of the typical sea drift garbage includes: size information, quantity information, and positioning longitude and latitude information.

Images