Intelligent identification method for floating objects in water area

Abstract

The invention relates to an intelligent identification method for floating objects in a water area, and the method comprises the following steps: S1, selecting a monitoring area, and setting three image sampling devices in the monitoring area; s2, respectively establishing a first image data set for the image sampling devices; s3, establishing a second image data set; s4, inputting a convolutional neural network for training to obtain a floating object identification model; s5, after real-time images are collected, if it is judged that the real-time images are floating objects in the floating object recognition model, an image sampling device with small illumination features in the same period is called, identification results of the images in the same period in the floating object recognition model are subjected to evidence judgment, if the identification results are consistent, the real-time images are judged to be floating objects, and if the identification results are not consistent, the other side is called, and a large number of judgment results are taken as the standard for output; and S6, after the real-time image is acquired, if the real-time image is judged to be abnormal in the floating object identification model, calling other image acquisition devices in sequence, and taking an output result which is finally called as a final result. The method has the advantages of small identification error and quick response.

Description

Technical Field

[0001] The invention relates to the technical field of water quality detection, and in particular to a method for intelligently identifying floating objects in water areas. Background Art

[0002] Water environment refers to the space where water is formed, distributed and transformed in nature. It refers to the natural environment surrounding human space and water bodies that can directly or indirectly affect human life and development, and the overall natural factors and related social factors of its normal function. Some also refer to the space where relatively stable natural water bodies with land as their boundaries are located.

[0003] At present, as the concept of environmental protection has been deeply rooted in people's hearts, pollution prevention and control work has been continuously carried out, especially in the field of water pollution prevention and control. How to quickly and in real time identify the pollution status of water areas is the key to water pollution prevention and control work, especially the identification and monitoring of floating objects, which can reduce the impact of floating objects on water bodies in a quick and clear way. The existing methods for identifying the status of floating objects in water areas mostly use manual detection, which is time-consuming and labor-intensive, and its timeliness is difficult to guarantee. In addition, the real-time video acquisition method is easily affected by the reflection of the water surface, resulting in misjudgment, and cannot provide efficient and accurate support for water pollution prevention and control work. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the technical problem to be solved by this patent application is how to provide an intelligent identification method for floating objects in water areas with small identification error and rapid response.

[0005] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0006] A method for intelligently identifying floating objects in waters comprises the following steps:

[0007] S1: Select a monitoring area and set three image sampling devices in the monitoring area, the angle between the three image sampling devices is 60 degrees, and the middle image sampling device is set facing the flow direction of the water area;

[0008] S2: establishing a first image data set for each image sampling device, wherein the first image data set includes clean water surface information, and generates illumination characteristics and weather characteristics according to the longitude and latitude of the selected area;

[0009] S3: establishing a second image data set, wherein the second image data set includes floating object features and water surface information with floating object features;

[0010] S4: inputting the first image data set and the second image data set in steps S2 and S3 into a convolutional neural network for training to obtain a floating object recognition model;

[0011] S5: Based on the sampling pattern of the middle image sampling device, after collecting the real-time image, if it is determined to be a floating object in the floating object recognition model, call the image sampling device with small illumination characteristics in the same period, and use the recognition result of its image in the floating object recognition model for supporting judgment. If they are consistent, it is determined to be a floating object. If they are inconsistent, call the image sampling device on the other side and output the judgment result with the larger number.

[0012] S6: Based on the sampling pattern of the middle image sampling device, after collecting the real-time image, if it is judged as abnormal in the floating object recognition model, call the image sampling device with small illumination characteristics in the same period, and use the recognition result of its image in the same period in the floating object recognition model to support the judgment. If it is judged as a floating object, the floating object is output. If the output result is not a floating object, call the image sampling device on the other side, and the result output by the last called image acquisition device is the final output result.

[0013] As an optimization, in step S1, the image sampling device is a video image collector of the same model, which can perform dynamic image acquisition and static image capture. At the same time, the image sampling device has image processing functions, including but not limited to image rotation, image reflection transformation, image flip transformation, image scaling transformation, image translation transformation, image scale transformation, image contrast transformation, image noise disturbance, and image color transformation.

[0014] As an optimization, in step S2, the illumination feature is that the image sampling device acquires the water surface reflection image information presented at different illumination angles in the area.

[0015] As an optimization, in step S2, the weather information is the water surface image information in rainy and snowy weather obtained by image sampling information in the area.

[0016] As an optimization, in step S3, the floating object information is input and called from an external database of commonly used floating objects in waters as training information. At the same time, a continuous input port is opened during actual use to enter the newly added floating object information that is misjudged or wrongly judged.

[0017] As an optimization, in step S3, the water surface information with floating object features includes combining the above floating object information with an image of clean water surface information, and enriching it by inputting images generated during continuous use.

[0018] As an optimization, in step S4, the floating object recognition model also includes displaying the coordinate position in the image area, which is presented on the framed floating object and displayed in the x, y form of a static image.

[0019] As an optimization, the convolutional neural network adopts a CNN neural network.

[0020] The present invention has the following advantages:

[0021] This solution can complete the autonomous identification of floating objects on the surface of flowing waters. Compared with the current video monitoring method, it adds a floating object recognition model. After the main image acquisition device obtains the floating object information, it calls the images of the same period at different angles for auxiliary judgment, which greatly improves the accuracy of judgment. In addition, this solution can realize autonomous monitoring and identification, has strong timeliness, can achieve uninterrupted real-time monitoring, and releases labor. Finally, this solution can supplement information manually or autonomously, and can independently supplement the materials that are not included, thereby improving the accuracy of judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The present invention is a flow chart of a method for intelligently identifying floating objects in water areas.

[0023] Figure 2 This is a distribution diagram of the image sampling devices of the method for intelligent identification of floating objects in water areas described in the present invention. DETAILED DESCRIPTION

[0024] The present invention is further described in detail below in conjunction with the accompanying drawings. In the description of the present invention, it should be understood that the directions or positional relationships indicated by directional words such as "upper, lower" and "top, bottom" are usually based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description. Unless otherwise stated, these directional words do not indicate or imply that the devices or components referred to must have a specific direction or be constructed and operated in a specific direction, and therefore cannot be understood as limiting the scope of protection of the present invention; the directional words "inside and outside" refer to the inside and outside relative to the contours of each component itself.

[0025] like Figure 1-2As shown, a method for intelligently identifying floating objects in waters comprises the following steps: S1: selecting a monitoring area, and setting three image sampling devices in the monitoring area, wherein the angles between the three image sampling devices are 60 degrees, and the image sampling device in the middle is set facing the flow direction of the waters; S2: establishing a first image data set for each image sampling device, wherein the first image data set contains clean water surface information, and at the same time, generates illumination features and weather features according to the latitude and longitude of the selected area; S3: establishing a second image data set, wherein the second image data set contains floating object features and water surface information with floating object features; S4: inputting the first image data set and the second image data set in steps S2 and S3 into a convolutional neural network for training to obtain a floating object recognition model; S5: taking the sampling pattern of the middle image sampling device as a reference, after collecting a real-time image, if it is determined to be a floating object in the floating object recognition model, calling an image sampling device with a small illumination feature in the same period, and using the recognition result of its image in the same period in the floating object recognition model for corroborative judgment, if they are consistent, it is determined to be a floating object, and if they are inconsistent, calling the image sampling device on the other side, and outputting the judgment result with the larger number as the standard. S6: Based on the sampling pattern of the middle image sampling device, after collecting the real-time image, if it is judged as abnormal in the floating object recognition model, call the image sampling device with small illumination characteristics in the same period, and use the recognition result of its image in the same period in the floating object recognition model to support the judgment. If it is judged as a floating object, the floating object is output. If the output result is not a floating object, call the image sampling device on the other side, and the result output by the last called image acquisition device is the final output result.

[0026] In this embodiment, in step S1, the image sampling device is a video image collector of the same model, which can perform dynamic image acquisition and static image capture. At the same time, the image sampling device has image processing functions, including but not limited to image rotation, image reflection transformation, image flip transformation, image scaling transformation, image translation transformation, image scale transformation, image contrast transformation, image noise disturbance, and image color transformation.

[0027] In this embodiment, in step S2, the illumination feature is the water surface reflection image information presented at different illumination angles acquired by the image sampling device in the area.

[0028] In this embodiment, in step S2, the weather information is water surface image information in rainy and snowy weather obtained by image sampling information in the area.

[0029] In this embodiment, in step S3, the floating object information is input and called from an external database of commonly used floating objects in waters as training information. At the same time, a continuous input port is opened during actual use to enter the newly added floating object information that is misjudged or wrongly judged.

[0030] In this embodiment, in step S3, the water surface information with floating object features includes combining the above floating object information with an image of clean water surface information, and enriching it by inputting images generated during continuous use.

[0031] In this embodiment, in step S4, the floating object recognition model also includes displaying the coordinate position in the image area, which is presented on the framed floating object and displayed in the form of x, y of a static image.

[0032] In this embodiment, the convolutional neural network adopts a CNN neural network.

[0033] The present invention has the following advantages:

[0034] This solution can complete the autonomous identification of floating objects on the surface of flowing waters. Compared with the current video monitoring method, it adds a floating object recognition model. After the main image acquisition device obtains the floating object information, it calls the images of the same period at different angles for auxiliary judgment, which greatly improves the accuracy of judgment. In addition, this solution can realize autonomous monitoring and identification, has strong timeliness, can achieve uninterrupted real-time monitoring, and releases labor. Finally, this solution can supplement information manually or autonomously, and can independently supplement the materials that are not included, thereby improving the accuracy of judgment.

[0035] Finally, it should be noted that those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for intelligently identifying floating objects in waters, characterized in that: The following steps are involved: S1: Select a monitoring area and set three image sampling devices in the monitoring area, the angle between the three image sampling devices is 60 degrees, and the middle image sampling device is set facing the flow direction of the water area; S2: establishing a first image data set for each image sampling device, wherein the first image data set includes clean water surface information, and generates illumination characteristics and weather characteristics according to the longitude and latitude of the selected area; S3: establishing a second image data set, wherein the second image data set includes floating object features and water surface information with floating object features; S4: inputting the first image data set and the second image data set in steps S2 and S3 into a convolutional neural network for training to obtain a floating object recognition model; S5: Based on the sampling pattern of the middle image sampling device, after collecting the real-time image, if it is determined to be a floating object in the floating object recognition model, call the image sampling device with small illumination characteristics in the same period, and use the recognition results of its image in the same period in the floating object recognition model for corroborative judgment. If they are consistent, it is determined to be a floating object. If they are inconsistent, call the image sampling device on the other side, and the judgment result with the larger number shall prevail for output; S6: Based on the sampling pattern of the middle image sampling device, after collecting the real-time image, if it is judged as abnormal in the floating object recognition model, call the image sampling device with small illumination characteristics in the same period, and use the recognition result of its image in the same period in the floating object recognition model to support the judgment. If it is judged as a floating object, the floating object is output. If the output result is not a floating object, call the image sampling device on the other side, and the result output by the last called image acquisition device is the final output result.

2. The method for intelligently identifying floating objects in waters according to claim 1, characterized in that: In step S1, the image sampling device is a video image collector of the same model, which can perform dynamic image acquisition and static image capture. At the same time, the image sampling device has image processing functions, including but not limited to image rotation, image reflection transformation, image flip transformation, image scaling transformation, image translation transformation, image scale transformation, image contrast transformation, image noise disturbance, and image color transformation.

3. The method for intelligently identifying floating objects in waters according to claim 2, characterized in that: In step S2, the illumination feature is the water surface reflection image information presented at different illumination angles acquired by the image sampling device in the area.

4. The method for intelligently identifying floating objects in waters according to claim 3, characterized in that: In step S2, the weather information is water surface image information in rainy and snowy weather obtained by image sampling information in the area.

5. The method for intelligently identifying floating objects in waters according to claim 4, characterized in that: In step S3, the floating object information is input and called from an external database of commonly used floating objects in waters as training information. At the same time, a continuous input port is opened during actual use to enter the newly added floating object information that is misjudged or wrongly judged.

6. A method for intelligently identifying floating objects in waters according to claim 5, characterized in that: In step S3, the water surface information with floating object features includes combining the above floating object information with the image of clean water surface information, and enriching it by inputting images generated during continuous use.

7. The method for intelligently identifying floating objects in waters according to claim 6, characterized in that: In step S4, the floating object recognition model also includes displaying the coordinate position in the image area, which is presented on the framed floating object in the form of x, y of a static image.

8. The method for intelligently identifying floating objects in waters according to claim 7, characterized in that: The convolutional neural network adopts a CNN neural network.

Images