CN-122023335-A - River channel ecological moisturizing effect evaluation method and system based on image detection

CN122023335ACN 122023335 ACN122023335 ACN 122023335ACN-122023335-A

Abstract

The application provides a river channel ecological moisturizing effect evaluation method and system based on image detection, and relates to the field of image detection, wherein the method comprises the steps of obtaining illumination intensity, water body nutrient content and tide cycle state information of an environment where a camera unit is positioned; according to illumination intensity, water nutrient content and tide cycle state information, the formation tendency of a biological film on the surface of a lens is estimated, in an image correction flow, the surface of the lens is subjected to exploratory imaging to obtain exploratory image data, an obtained quantized result is obtained, correction parameters for compensating influence are generated according to the quantized result, adjustment of brightness, color and definition is carried out on an ecological image acquired subsequently, the ecological condition is judged based on the adjusted ecological image, the problem that the accuracy of an estimation result is affected due to distortion of image data caused by lens dirt in the prior art is effectively solved, and the problems of image blurring, color distortion or uneven brightness and the like caused by a dirt layer are avoided.

Inventors

YANG JIAQIAN
WU BAICAI
CHEN LU
Lu Beiyang
YAN CHENHAO
LUO JIE
YU MIN

Assignees

浙江问源环保科技股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260130

Claims (10)

1. The river channel ecological moisturizing effect evaluation method based on image detection is characterized by comprising the following steps of: acquiring illumination intensity, water nutrient content and tide cycle state information of an environment where a camera unit is located; According to the illumination intensity, the water nutrient content and the tide cycle state information, evaluating the formation tendency of the biological film on the surface of the lens to obtain an evaluation result of the formation tendency; When the evaluation result of the formation tendency reaches a preset condition, starting an image correction flow; in the image correction flow, detecting imaging is carried out on the surface of the lens so as to obtain image data after light rays pass through the dirt layer, and detecting image data is obtained; Analyzing the difference between the detected image data and the reference image in a clean state to quantify the scattering, deflection or absorption influence of the dirt layer on light rays, and obtaining a quantification result of the scattering, deflection or absorption influence; Generating correction parameters for compensating the influence according to the quantification result of the scattering, deflection or absorption influence; applying the correction parameters to the subsequently acquired ecological images, and adjusting brightness, color and definition; and judging the ecological condition based on the adjusted ecological image.
2. The method for evaluating the ecological water replenishing effect of a river channel based on image detection according to claim 1, wherein the step of analyzing the difference between the detected image data and the reference image in the clean state to quantify the scattering, deflection or absorption effect of the dirt layer on the light, and obtaining the quantified result of the scattering, deflection or absorption effect comprises the steps of: setting a reference panel inside the camera unit; Driving an auxiliary light source to illuminate the reference panel; capturing an image of the reference panel when illuminated; comparing the image of the reference panel with a reference image shot by the light source in an initial state to obtain the output characteristic of the auxiliary light source; Generating a theoretical cleaning pattern image by utilizing the output characteristic of the auxiliary light source and combining an external calibration pattern integrated on the surface of the lens; and carrying out differential analysis on the actually captured external calibration pattern image influenced by the dirt and the theoretical cleaning pattern image so as to quantify the scattering, deflection or absorption influence of the dirt layer on the light, and obtaining a quantification result of the scattering, deflection or absorption influence.
3. The method for evaluating the ecological water replenishing effect of a river channel based on image detection according to claim 1, wherein the step of evaluating the formation tendency of the biological film on the surface of the lens according to the illumination intensity, the water body nutrient content and the tide cycle state information, and obtaining the evaluation result of the formation tendency comprises the following steps: acquiring temperature information of a local water body at a preset position of a camera lens; acquiring dynamic characteristics of microscopic water flow on the surface of the lens; evaluating the formation tendency of the biological film on the surface of the lens according to the illumination intensity, the water nutrient content, the tide cycle state information, the local water temperature information and the microscopic water flow dynamics characteristics; When the local water body temperature information exceeds a preset threshold value and the microscopic water flow dynamics feature shows that a water flow retention or vortex area exists, the local water body temperature information and the microscopic water flow dynamics feature are used as weighting factors for evaluating the formation tendency of the biological film so as to adjust the evaluation result of the formation tendency of the biological film on the surface of the lens.
4. The method for evaluating the ecological water replenishing effect of a river channel based on image detection according to claim 1, wherein the step of evaluating the formation tendency of the biological film on the surface of the lens according to the illumination intensity, the water body nutrient content and the tide cycle state information, and obtaining the evaluation result of the formation tendency comprises the following steps: acquiring type information of dominant microbial communities in the water body; Identifying the type of a biological film formed on the surface of the lens in the current environment according to the illumination intensity, the water nutrient content, the tide cycle state information and the dominant microorganism community type information; invoking prestored optical influence characteristic data specific to the type of the biological film; and according to the identified type of the biological film and the characteristic data of the special optical influence, adjusting the evaluation result of the formation tendency of the biological film on the surface of the lens by combining the illumination intensity, the water body nutrient content and the tide cycle state information.
5. The method for evaluating the ecological water replenishing effect of a river channel based on image detection according to claim 1, wherein the step of obtaining the illumination intensity, the water nutrient content and the tide cycle state information of the environment in which the camera unit is located comprises the steps of: disposing a plurality of sensors at preset positions of the camera unit to form a sensor array for acquiring illumination intensity, water nutrient content and tide cycle state information; Periodically driving a mechanical cleaning unit to physically clean each sensor surface of the sensor array; before and after each cleaning, driving the sensor array to perform one self-checking measurement to acquire measurement data before cleaning and measurement data after cleaning; Comparing the post-cleaning measurement data with the pre-cleaning measurement data to evaluate a measurement bias of each sensor; when the measurement deviation of a certain sensor exceeds a preset threshold value, marking the data of the certain sensor as abnormal, and removing the data from the sensor array to obtain the residual sensor data; and acquiring the illumination intensity, the water body nutrient content and the tide cycle state information by using the residual sensor data in the sensor array through data fusion processing.
6. The method for evaluating the ecological water replenishing effect of a river channel based on image detection according to claim 1, wherein the step of starting the image correction flow when the evaluation result of the formation tendency reaches a preset condition comprises the steps of: When the evaluation result of the formation tendency reaches a preset condition, acquiring water level information at the current moment; judging whether the water level information is in a preset complete immersion interval of the lens; When the water level information is in the preset complete immersion interval of the lens, starting the image correction flow; And when the water level information is not in the preset lens complete immersion interval, waiting until the water level information is in the preset lens complete immersion interval, and starting the image correction flow.
7. The method for evaluating the ecological water replenishing effect of a river channel based on image detection according to claim 1, wherein in the image correction process, the step of performing exploratory imaging on the lens surface to obtain image data after light passes through a dirt layer, and obtaining exploratory image data comprises the following steps: driving a camera unit to continuously acquire a multi-frame image sequence of the lens surface at a preset frame rate; Performing inter-frame difference analysis on the multi-frame image sequence, and identifying and removing image frames with instantaneous high-intensity light spots or fast moving fuzzy areas to obtain residual image frames; And carrying out image superposition processing on the rest image frames to generate image data of the light rays passing through the dirt layer, so as to obtain detectable image data.
8. The method for evaluating the ecological moisturizing effect of a river course based on image detection according to claim 7, wherein the step of performing the inter-frame difference analysis on the multi-frame image sequence, identifying and rejecting the image frames with the instantaneous high-intensity light spots or the fast moving blurred regions, and obtaining the remaining image frames comprises the steps of: acquiring each frame of image in the multi-frame image sequence; Detecting characteristic points of each frame of image; Calculating a motion vector between adjacent image frames based on the feature points; performing global motion compensation on the image sequence according to the motion vector so as to eliminate integral displacement or deformation of the camera unit and obtain an image sequence subjected to global motion compensation; Performing inter-frame difference analysis on the global motion compensated image sequence; And identifying and removing the image frames with the instantaneous high-intensity light spots or the fast moving blur areas to obtain the rest image frames.
9. The method for evaluating the ecological water replenishing effect of a river channel based on image detection according to claim 7, wherein the step of performing image superposition processing on the remaining image frames to generate image data after the light passes through the dirt layer, and obtaining the detected image data comprises the following steps: acquiring background turbidity information of the current water body; performing background scattering compensation processing on the rest image frames according to the background turbidity information; And performing image superposition processing on the image frame subjected to the background scattering compensation processing to generate image data of the light rays passing through the dirt layer, thereby obtaining detectable image data.
10. River channel ecological moisturizing effect evaluation system based on image detection, which is characterized by comprising: The information acquisition module is used for acquiring the illumination intensity of the environment where the camera unit is located, the content of nutrient substances in the water body and the tidal cycle state information; The trend evaluation module is used for evaluating the formation trend of the biological film on the surface of the lens according to the illumination intensity, the water body nutrient content and the tide cycle state information to obtain an evaluation result of the formation trend; the flow starting module is used for starting an image correction flow when the evaluation result of the formation tendency reaches a preset condition; The detection imaging module is used for carrying out detection imaging on the surface of the lens in the image correction flow so as to obtain image data of light rays passing through the dirt layer and obtain detection image data; The difference analysis module is used for analyzing the difference between the detected image data and the reference image in a clean state so as to quantify the scattering, deflection or absorption influence of the dirt layer on the light and obtain a quantification result of the scattering, deflection or absorption influence; A parameter generation module for generating correction parameters for compensating the influence according to the quantification result of the scattering, deflection or absorption influence; the image adjusting module is used for applying the correction parameters to the subsequently acquired ecological image and adjusting brightness, color and definition; And the condition judging module is used for judging the ecological condition based on the adjusted ecological image.

Description

River channel ecological moisturizing effect evaluation method and system based on image detection Technical Field The application relates to the technical field of image detection, in particular to a river ecological water replenishing effect evaluation method and system based on image detection. Background In water environments, particularly open waters such as riverways, lakes, or oceans, the use of image detection techniques to monitor water quality, vegetation growth, or biological activity has become an important tool. However, the camera device disposed near the water body for a long time often has the lens surface affected by various environmental factors, which causes problems such as blurring, color distortion or uneven brightness of the image, so that the accuracy of the image data is seriously affected, and the subsequent image analysis and evaluation results are not reliable. The particularity of tidal-type river environments presents unique challenges for the long-term stable operation of such image detection systems. Over time, these microorganisms, mainly algae and bacteria of various types, that adhere to the lens surface can start to multiply in large numbers. The extracellular polymers secreted by these microorganisms are capable of tightly adhering microbial cells to the lens surface and forming a structured, highly adherent biofilm. Further, the absorbed silt, organic debris and biological film are mixed together, and a non-uniform, semitransparent and abnormally firm composite dirt layer can be formed through sunlight irradiation and air drying. The composite dirt layer formed by mixing the biological film and the physical particles has obvious interference on light entering the camera lens. When these locally quality compromised images are transmitted into an image analysis system, their internal logic processing modules can face significant challenges. Because the local brightness of the image is reduced or the color of the image is deviated due to the composite dirt layer on the lens, even if the river water is actually clear, the system can also falsely judge that the turbidity of the water quality is increased because the gray value in the received image data is abnormally increased or the color deviates from the normal water color range, so that the coverage area of vegetation or the growth vigor of the vegetation is falsely underestimated by the system. The degradation of optical performance of the lens caused by the unique biophysical process in the tidal river environment is transmitted to the logic processing module layer by layer through the image sensing link, and finally, the systematic deviation of the evaluation result is caused. Disclosure of Invention The application discloses a river channel ecological moisturizing effect evaluation method and system based on image detection, and aims to solve the problems that in a water body environment, a dirt layer on the surface of a camera lens has a serious influence on image data accuracy, so that subsequent image analysis and evaluation results are unreliable, and particularly in a tidal river channel environment, a dirt forming mechanism is complex, and higher requirements are put forward on the stability and evaluation accuracy of a monitoring system. The technical scheme of the application is as follows: The application discloses a river channel ecological moisturizing effect evaluation method based on image detection, which comprises the following steps of obtaining illumination intensity, water nutrient content and tide cycle state information of an environment where a camera unit is located, evaluating formation tendency of a lens surface biological film according to the illumination intensity, the water nutrient content and the tide cycle state information to obtain an evaluation result of the formation tendency, starting an image correction flow when the evaluation result of the formation tendency reaches a preset condition, performing exploratory imaging on the lens surface in the image correction flow to obtain image data after light passes through a dirt layer to obtain exploratory image data, analyzing differences of the exploratory image data and a reference image in a clean state to quantify scattering, deflection or absorption effects of the dirt layer on the light to obtain quantification results of the scattering, deflection or absorption effects, and generating correction parameters for compensating the effects according to the quantification results of the scattering, deflection or absorption effects; Applying correction parameters to the subsequently acquired ecological images, and adjusting brightness, color and definition; And judging the ecological condition based on the adjusted ecological image. In a second aspect, the application also discloses a river ecological water replenishing effect evaluation system based on image detection, which comprises: The information acquisition module is used for acqui