CN111833369A - Alum image processing method, system, medium and electronic device - Google Patents
Alum image processing method, system, medium and electronic device Download PDFInfo
- Publication number
- CN111833369A CN111833369A CN202010703199.XA CN202010703199A CN111833369A CN 111833369 A CN111833369 A CN 111833369A CN 202010703199 A CN202010703199 A CN 202010703199A CN 111833369 A CN111833369 A CN 111833369A
- Authority
- CN
- China
- Prior art keywords
- image
- alumen ustum
- alum blossom
- target object
- alum
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 229940037003 alum Drugs 0.000 title claims abstract description 120
- 238000003672 processing method Methods 0.000 title claims description 12
- 238000003709 image segmentation Methods 0.000 claims abstract description 63
- 238000000034 method Methods 0.000 claims abstract description 47
- 238000012545 processing Methods 0.000 claims abstract description 41
- 238000012549 training Methods 0.000 claims abstract description 18
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 16
- 230000008569 process Effects 0.000 claims abstract description 13
- 238000004590 computer program Methods 0.000 claims description 13
- 230000008602 contraction Effects 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 12
- 230000011218 segmentation Effects 0.000 claims description 10
- 238000011176 pooling Methods 0.000 claims description 8
- 230000004913 activation Effects 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 7
- 238000005070 sampling Methods 0.000 claims description 6
- 238000013461 design Methods 0.000 claims description 3
- 239000013598 vector Substances 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims description 2
- 230000008859 change Effects 0.000 abstract description 6
- 230000015271 coagulation Effects 0.000 abstract description 4
- 238000005345 coagulation Methods 0.000 abstract description 4
- 238000004458 analytical method Methods 0.000 abstract 1
- 238000012544 monitoring process Methods 0.000 abstract 1
- 238000004891 communication Methods 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000001514 detection method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000010606 normalization Methods 0.000 description 4
- 238000005286 illumination Methods 0.000 description 3
- 230000003213 activating effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 239000000701 coagulant Substances 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 239000013043 chemical agent Substances 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000005189 flocculation Methods 0.000 description 1
- 230000016615 flocculation Effects 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 239000003595 mist Substances 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Geometry (AREA)
- Image Processing (AREA)
Abstract
The invention provides a method, a system, a medium and an electronic device for processing alum blossom images, wherein the method comprises the following steps: acquiring underwater alum blossom image sample data, and performing data annotation; establishing an alum blossom image segmentation model according to the marked data, and training; acquiring a real-time image of a target object, inputting the real-time image into a trained alumen ustum image segmentation model for image segmentation, and acquiring alumen ustum contour information of the target object; acquiring alum blossom position information of a target object according to the alum blossom outline information; according to the method, the position and contour information of the alum blossom is obtained through image acquisition and image segmentation processing, the real-time monitoring and analysis of alum blossom characteristics in the coagulation process are replaced by naked eyes, manual participation is not needed in the judgment process, the problems that the dosing amount is controlled according to the change of factors such as water quality and the like and dosing cannot be measured are solved, the reliability and the precision are improved, and the system error is reduced.
Description
Technical Field
The invention relates to the field of industrial detection, in particular to an alum blossom image processing method, an alum blossom image processing system, an alum blossom image processing medium and electronic equipment.
Background
Coagulation is a process of gathering colloidal particles and micro suspended matters in water by a certain method (such as adding chemical agents), is an important link of water treatment, and not only influences the subsequent treatment process, but also influences the effluent quality and the treatment cost. How to determine the better coagulant adding amount according to the change of the water quality of the incoming water by a dosing system is a problem which is generally concerned and needs to be solved in the water supply and drainage industry for a long time.
At present, the existing judgment mode is to manually judge whether the dosage needs to be adjusted or not according to the alum blossom characteristics identified by naked eyes in the coagulation process. However, the manual judgment method cannot control the dosage on line in real time according to the change of factors such as the quality of raw water and the like, and cannot achieve the real metering dosage, so that the water quality condition is difficult to master, and the coagulant is excessively consumed and labor is wasted.
Disclosure of Invention
In view of the above-mentioned shortcomings of the prior art, the present invention provides a method, a system, a medium and an electronic device for processing alum blossom image, so as to solve the above-mentioned technical problems.
The invention provides an alum blossom image processing method, which comprises the following steps:
acquiring underwater alum blossom image sample data, and performing data annotation;
establishing an alum blossom image segmentation model according to the marked data, and training;
acquiring a real-time image of a target object, inputting the real-time image into a trained alumen ustum image segmentation model for image segmentation, and acquiring alumen ustum contour information of the target object;
and acquiring the position information of the alum flocs of the target object according to the alum floc outline information.
Optionally, the alumen ustum outline information includes an alumen ustum outline region, a region area, an alumen ustum image gray variance, an alumen ustum image gradient, an alumen ustum image kurtosis, an alumen ustum image entropy, and an alumen ustum image fractal dimension.
Optionally, underwater alum blossom image sample data is obtained and subjected to data annotation to form an alum blossom image data set, and image preprocessing is performed on the alum blossom image data set, where the preprocessing includes image defogging processing, and the image defogging processing includes obtaining a dark channel image by obtaining a minimum value in RGB components of each pixel in an image.
Optionally, the dark channel image is obtained by the following formula:
wherein, JdarkFor each channel of the dark channel image; j. the design is a squarecFor each channel of the color image; Ω (x) is a window centered on pixel x.
Optionally, the preprocessing further includes a low-light image enhancement processing, and the low-light image enhancement processing includes
Decomposing the image into an incident image and a reflected image;
and removing the incident attribute in the image and keeping the original image attribute.
Optionally, a boundary rectangular frame of the contour of the target object is obtained according to the alumen ustum contour information, and alumen ustum position information is obtained according to coordinates of the boundary rectangular frame.
Optionally, an alum blossom outline region and a region area of the target object are determined according to the bounding rectangle frame, a water background region is determined according to the alum blossom outline region, and image segmentation is performed on alum blossom and a water background in the target object image.
Optionally, the alumen ustum image segmentation model comprises
A contraction path for obtaining context information in the target object;
and the extended path is used for positioning the part needing to be segmented in the target object.
Optionally, a maximum pooling layer is connected behind every two convolution layers in the contraction path, and an original image of the target object is down-sampled by an activation function behind each convolution layer, so that the number of channels is doubled during each down-sampling.
Optionally, the extension path includes a plurality of convolution layers for performing upsampling at each step, and each step of upsampling is added with a feature map corresponding to the contraction path, respectively, and the last layer of the alumen ustum image segmentation model is the convolution layer for converting the feature vectors of the channels into the number of required classification results.
Optionally, the pixel-level features in the image are expanded, and image segmentation is performed by combining local clues and global clues, where the expansion includes fusing multiple different pyramid scale features, using convolution to reduce a level channel for each level, obtaining the size of the image before pooling through a bilinear difference, and finally merging the size together.
The invention also provides an alum blossom image processing system, which comprises:
the data sample module is used for acquiring underwater alum blossom image sample data and carrying out data annotation;
the image processing model is used for establishing an alum blossom image segmentation model according to the marked data and training;
the image acquisition module is used for acquiring a real-time image of the target object and inputting the real-time image into the trained alum blossom image segmentation model for image segmentation;
the alumen ustum image segmentation model acquires alumen ustum contour information of a target object, and acquires alumen ustum position information of the target object according to the alumen ustum contour information.
Optionally, the alumen ustum image segmentation model includes symmetric models:
a contraction path for obtaining context information in the target object;
and the extended path is used for positioning the part needing to be segmented in the target object.
Optionally, a maximum pooling layer is connected behind every two convolution layers in the contraction path, and an original image of the target object is down-sampled by an activation function behind each convolution layer, and the number of channels is doubled during each down-sampling.
Optionally, the expansion path includes a plurality of convolution layers for performing upsampling in each step, and each step of upsampling is added to the feature map corresponding to the contraction path.
Optionally, the alumen ustum image segmentation model generates a prediction segmentation result of the verification picture after each batch of training is completed, and trains model parameters according to the labeling data and the prediction segmentation result, and determines the final recognition accuracy.
Optionally, according to the image segmentation result of the alum blossom image segmentation model, the model with the highest accuracy is selected as the optimal model.
The invention also provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method of any one of the above.
The present invention also provides an electronic terminal, comprising: a processor and a memory; the memory is adapted to store a computer program and the processor is adapted to execute the computer program stored by the memory to cause the terminal to perform the method as defined in any one of the above.
The invention has the beneficial effects that: according to the alumen ustum image processing method, the alumen ustum image processing system, the electronic equipment and the medium, the position and contour information of the alumen ustum is obtained through image acquisition and image segmentation processing, the characteristics of the alumen ustum in the coagulation process are monitored and analyzed in real time instead of naked eyes, manual participation is not needed in the judgment process, the problems that the dosing amount is controlled according to the change of factors such as water quality and the like and dosing cannot be measured are solved, the reliability and the precision are improved, and the system error is reduced.
Drawings
FIG. 1 is a schematic flow chart of a alum blossom image processing method according to an embodiment of the present invention.
FIG. 2 is a schematic diagram of a network structure of a segmentation model of an alumen ustum image in the method for processing an alumen ustum image according to the embodiment of the present invention.
FIG. 3 is a schematic diagram of an activation function in the alumen ustum image processing method according to the embodiment of the present invention.
FIG. 4 is a schematic diagram of acquiring real-time images of underwater alum floc in the alum floc image processing method according to the embodiment of the present invention.
FIG. 5 is a schematic view of an outline region of alum blossom in the alum blossom image processing method according to the embodiment of the present invention.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention, and the components related to the present invention are only shown in the drawings rather than drawn according to the number, shape and size of the components in actual implementation, and the type, quantity and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated.
In the following description, numerous details are set forth to provide a more thorough explanation of embodiments of the present invention, however, it will be apparent to one skilled in the art that embodiments of the present invention may be practiced without these specific details, and in other embodiments, well-known structures and devices are shown in block diagram form, rather than in detail, in order to avoid obscuring embodiments of the present invention.
As shown in fig. 1, the alumen ustum image processing method in the present embodiment includes:
acquiring underwater alum blossom image sample data, and performing data annotation;
establishing an alum blossom image segmentation model according to the marked data, and training;
acquiring a real-time image of a target object, inputting the real-time image into a trained alumen ustum image segmentation model for image segmentation, and acquiring alumen ustum contour information of the target object;
and acquiring the position information of the alum flocs of the target object according to the alum floc outline information.
In this embodiment, image data annotation needs to be performed on the obtained underwater clear image of alum blossom, that is, an image corresponding to the clear image and obtained by segmenting an alum blossom outline is generated, as shown in fig. 4, an image data set is formed, and optionally, the image data set can be divided into a test data set and a training data set according to a ratio of 1: 9. After the image data set is obtained, the image data set needs to be preprocessed, and the preprocessing mainly comprises image defogging processing, low-illumination image enhancement processing and normalization processing. And carrying out image normalization processing on the alum blossom image data set to normalize the gray value of the picture from 2 to 255 to 0 to 1. The image normalization in this embodiment may adopt a maximum and minimum normalization method, and the obtaining method is as follows:
where xi denotes the image pixel point values max (x), and min (x) denotes the maximum and minimum values of the image pixels, respectively.
In this embodiment, the image is defogged, and the minimum value of RGB components of each pixel in the original image is obtained to obtain a dark channel image, where the mathematical formula is:
wherein, JdarkEach channel representing a dark channel image; j. the design is a squarecRepresenting a colour imageEach channel; Ω (x) represents a window centered on pixel x;
carrying out minimum value filtering processing on the dark channel image, acquiring the front 0.1% pixel position according to the brightness of the dark channel image, acquiring an original image pixel value of a corresponding point corresponding to the original image, and recovering the dark channel image to obtain a defogged image, wherein the mathematical formula is as follows:
wherein J (x) is a defogged image under the dark channel algorithm, I (x) is an input original image, A is atmospheric illumination, namely the pixel value of the original image corresponding to 0.1% of the pixel position before the brightness of the image passing through the dark channel, t (x) is the transmissivity, t (x) is the brightness of the image passing through the dark channel, and0the threshold value set for preventing the image from being excessive to the white field is generally t0Standard calculation is 0.1, and the mathematical formula t (x) is:
where ω is a parameter for controlling the specific gravity of the remaining mist introduced so as not to distort the image, and ω 0.95 is an optimum value obtained by an experiment.
In this embodiment, an image S (x, y) can be decomposed into two different images, wherein the mathematical formula of the relationship is:
s (x, y) ═ R (x, y) × L (x, y) formula (5)
Wherein, R (x, y) is a reflection image, L (x, y) is an incident image, and a mathematical formula of an image R (x, y) finally formed in human eyes is as follows:
the low-illumination image enhancement processing firstly removes incident attributes L (x, y) in an image, and retains original image attributes S (x, y), and the mathematical formula is as follows:
wherein,
the operation represented is a convolution operation, K is the number of Gaussian surrounding functions, usually the value of K takes 3, and
f (x, y) is the center-surround function, and is given by:
wherein c is a gaussian surround scale, λ is a scale, and its value must satisfy the expression:
: | (x, y) dxdy ═ 1 formula (9)
After the final formed image is obtained, the pixel values are quantized to the range of 0 to 255 as the final output, and the mathematical formula is as follows:
in this embodiment, the alumen ustum image segmentation model is an underwater alumen ustum image segmentation neural network based on deep learning, and the input is a single enhanced alumen ustum image in the scene, and the image segmentation can be performed by learning the characteristics of the alumen ustum by using a U-net deep learning neural network, or by replacing the U-net with other image segmentation networks, such as SegNet, ENet, PSPNet and parset. Training an alum blossom image segmentation model: and (3) training an alum blossom image segmentation network in a supervision way, learning corresponding underwater alum blossom image characteristics, and outputting an image segmentation result. The alumen ustum image segmentation model in the embodiment mainly comprises two parts: a contracting path (contracting path) and an expanding path (expanding path). The contraction path is mainly used to capture context information (context information) in the picture, and the symmetrical expansion path is used to precisely locate the portion of the picture that needs to be segmented. As shown in fig. 2, in the present embodiment, optionally, 2 × 2 maximum pooling layers with a step size of 2 are followed by every two 3 × 3 convolutional layers (unpadded convolutional layers) in the systolic path, and each convolutional layer is followed by a ReLU activation function to perform downsampling operation on the original picture, in addition, each downsampling operation is increased by one time of channels. In the upsampling (deconstruction) of the extended path, there will be one 2 × 2 convolutional layer (the activation function is also ReLU) and two 3 × 3 convolutional layers at each step, while the upsampling of each step will add the feature map from the corresponding contracted path (clipped to keep the same shape). At the last layer of the network is a 1 x 1 convolutional layer, by which the 64-channel feature vectors can be converted into the required number of classification results (e.g., 2), and finally, the whole network has 23 convolutional layers. The alumen ustum image segmentation model in the embodiment can basically perform convolution operation on pictures with arbitrary shapes and sizes, especially on pictures with arbitrary sizes.
In this embodiment, the alumen ustum image segmentation model may use a PSPNet image segmentation network to perform image segmentation by combining local and global cues. The alum blossom image segmentation model reduces a grade channel by fusing four different pyramid scale features, using 1 × 1 convolution for each grade, obtaining the size before pooling through a bilinear difference value, and finally combining the sizes together, taking the sizes of the combined convolution kernels as 1 × 1, 3 × 3 and 5 × 5 as an example, and the mathematical expression is as follows:
wherein k is a convolution kernel, I refers to the number of pictures, and I is the ith picture.
In this embodiment, the alumen ustum image segmentation network is trained, supervised training is performed on the alumen ustum image segmentation network, corresponding underwater alumen ustum image features are learned, and an image segmentation result is output. Optionally, a series of data enhancement can be performed on the training image, and the steel bar picture in the scene is respectively cut, flipped, rotated, changed in brightness, contrast and saturation. In the image training process, the activating function uses a ReLU function, and the mathematical formula and the image of the activating function are shown in fig. 3, wherein the ReLU mathematical formula is as follows: and in order to avoid overfitting, the network learning rate is set by an exponential decay method, and a dropout regularization method is adopted, so that in each iteration of training the neural network, some neurons are randomly closed, and the complexity of the neural network is reduced.
In this embodiment, in the image segmentation process, a prediction segmentation result of a verification picture is generated after each batch of training is completed, and not only model parameters are trained according to the labeled picture and the prediction segmentation result, but also the final recognition accuracy is determined according to the labeled picture and the prediction segmentation result. When the training network carries out multiple iterations, the prediction segmentation result is converged towards the direction of the error of the labeled picture and the prediction segmentation result continuously, and then parameters are updated to each layer according to a chain rule through back propagation. And each iteration reduces propagation errors as much as possible according to the optimization direction of gradient descent, and finally obtains the final image segmentation result of all alum blossom images in the data set.
In this embodiment, the optimal model with the highest accuracy is stored based on the image segmentation result of the final alumen ustum image. In actual operation, firstly, a picture is obtained in real time through a camera, a single alum blossom image in the scene is used as input, the model automatically processes the image, the alum blossom characteristics are identified, prediction is carried out, and finally the image segmentation result of the alum blossom image is output. Identifying a highlight part of the binary image according to an image segmentation result to obtain a detection target; extracting the outermost contour of the detection target, and acquiring contour information, wherein the contour information comprises one of the following: the method comprises the following steps of (1) determining an alum blossom outline region, a region area, an alum blossom image gray level variance, an alum blossom image gradient, an alum blossom image kurtosis, an alum blossom image entropy and an alum blossom image fractal dimension; and acquiring a boundary rectangular frame of the outline to acquire position information. Determining a alum blossom outline region and a region area by extracting the alum blossom outline; the water background area is determined according to the alumen ustum outline area, as shown in fig. 3.
In the present embodiment, the area of the region is obtained by the following formula:
the method comprises the steps of obtaining equivalent diameters of all alum blossom outline areas, obtaining an average value, determining the equivalent diameters of all alum blossom outline areas, and obtaining a total surface of all alum blossom outline areas, wherein the equivalent diameters of all alum blossom outline areas are equal to the equivalent diameters of all alum blossom outline areas, and ContourARea is the total surface of all alum blossom outline areas.
After the image information of the segmented image is obtained, the area, the perimeter and the equivalent diameter of the region are calculated by analyzing and processing the related parameters, and the data can be used as the visual representation of the underwater alum blossom; the total area, the black area ratio, the average area of the alum flowers, the average circumference of the alum flowers and the average equivalent diameter of the alum flowers can be obtained, and the total distribution and the form condition of the current underwater alum flowers can be directly, clearly and accurately known through the calculation result; and further obtaining the grey variance of the alum blossom image, the gradient of the alum blossom image, the kurtosis of the alum blossom image, the entropy of the alum blossom image and the fractal dimension of the alum blossom image. In the present embodiment, the degree of dispersion of the random variable around the central value is represented by the acquired alumen ustum image gray variance, which is acquired by the following formula:
wherein s is the grey variance of the alum blossom image, xi is the grey of each pixel point of the alum blossom image, n is more than or equal to i and more than or equal to 1, M is the mean value of the grey of the alum blossom image, and n is the total number of the pixels of the alum blossom image.
In this embodiment, the gradient of the alumen ustum image indicates the degree of asymmetry between the random variable and the central distribution, and the gradient is rightward with a positive value and leftward with a negative value. The gradient of the alum blossom image is obtained by the following formula:
wherein, skew is the gradient of the alum blossom image, xi is the gray level of each pixel point of the alum blossom image, n is more than or equal to i and more than or equal to 1, M is the average gray level of the alum blossom image, and n is the total number of pixels of the alum blossom image.
Characterizing the characteristic number of the peak value height of a probability density distribution curve at the average value by the kurtosis of the alum blossom image, wherein the kurtosis of the alum blossom image is expressed by the following mathematic expression:
wherein kurt is the kurtosis of the alumen ustum image, xi is the gray level of each pixel point of the alumen ustum image, n is more than or equal to i and more than or equal to 1, M is the average gray level of the alumen ustum image, and n is the total number of pixels of the alumen ustum image.
The entropy value is image entropy, namely information entropy of an image, and the degree of disorder of information is represented by the magnitude of the entropy. Generally, the amount of information included in an image is large, the larger the entropy value is, the entropy of the alum blossom image is obtained by the following formula:
wherein H is the entropy of the alumen ustum image, and P (i) is the ratio of the number of pixels with the pixel value of i to the total number of pixels of the alumen ustum image
In this embodiment, the fractal dimension of the alum blossom image includes information such as density of alum blossom, and can be used to express flocculation effect, and the change of the fractal dimension can reflect the formation process and the rule of alum blossom, and the fractal dimension of the alum blossom image in this embodiment is obtained by the following formula:
A＝＝αLDfformula (17)
Wherein A is the area of the alumen ustum outline region, L is the outline perimeter of the alumen ustum outline region, alpha is the proportionality constant of the alumen ustum outline region, and Df is the fractal dimension of the alumen ustum image.
lnA ═ DflnL + ln alpha type (18)
And acquiring the areas A and the perimeters P corresponding to all alumen ustum, respectively carrying out logarithm removal, and fitting the data by using a least square method to obtain a corresponding straight line, wherein the slope of the straight line is the fractal dimension in the processing time period.
Correspondingly, the present embodiment further provides an alum blossom image processing system, including:
the data sample module is used for acquiring underwater alum blossom image sample data and carrying out data annotation;
the image processing model is used for establishing an alum blossom image segmentation model according to the marked data and training;
the image acquisition module is used for acquiring a real-time image of the target object and inputting the real-time image into the trained alum blossom image segmentation model for image segmentation;
the method comprises the steps of obtaining alum blossom outline information of a target object by an alum blossom image segmentation model, and obtaining alum blossom position information of the target object according to the alum blossom outline information.
In the embodiment, an object frame is determined through an image processing model, an alum blossom outline is extracted, and an alum blossom outline region and a region area are determined; the water background area is determined according to the alumen ustum outline area, the target underwater alumen ustum and the water background are divided, the picture is concise and clear, the field operator can know the real-time situation of the underwater alumen ustum more intuitively, and the basis and reference are provided for subsequent detection and regulation. The image processing process is as described above.
The present embodiment also provides a computer-readable storage medium on which a computer program is stored, which when executed by a processor implements any of the methods in the present embodiments.
The present embodiment further provides an electronic terminal, including: a processor and a memory;
the memory is used for storing computer programs, and the processor is used for executing the computer programs stored by the memory so as to enable the terminal to execute the method in the embodiment.
The computer-readable storage medium in the present embodiment can be understood by those skilled in the art as follows: all or part of the steps for implementing the above method embodiments may be performed by hardware associated with a computer program. The aforementioned computer program may be stored in a computer readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
The electronic terminal provided by the embodiment comprises a processor, a memory, a transceiver and a communication interface, wherein the memory and the communication interface are connected with the processor and the transceiver and are used for completing mutual communication, the memory is used for storing a computer program, the communication interface is used for carrying out communication, and the processor and the transceiver are used for operating the computer program so that the electronic terminal can execute the steps of the method.
In this embodiment, the Memory may include a Random Access Memory (RAM), and may also include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory.
The Processor may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the device can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, or a discrete hardware component.
In the above-described embodiments, reference in the specification to "the present embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least some embodiments, but not necessarily all embodiments. The multiple occurrences of "the present embodiment" do not necessarily all refer to the same embodiment.
In the embodiments described above, although the present invention has been described in conjunction with specific embodiments thereof, many alternatives, modifications, and variations of these embodiments will be apparent to those skilled in the art in light of the foregoing description. For example, other memory structures (e.g., dynamic ram (dram)) may use the discussed embodiments. The embodiments of the invention are intended to embrace all such alternatives, modifications and variances that fall within the broad scope of the appended claims.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
The invention is operational with numerous general purpose or special purpose computing system environments or configurations. For example: personal computers, server computers, hand-held or portable devices, tablet-type devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, and the like.
The invention may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
The foregoing embodiments are merely illustrative of the principles of the present invention and its efficacy, and are not to be construed as limiting the invention. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.
Claims (19)
1. An alum blossom image processing method is characterized by comprising the following steps:
acquiring underwater alum blossom image sample data, and performing data annotation;
establishing an alum blossom image segmentation model according to the marked data, and training;
acquiring a real-time image of a target object, inputting the real-time image into a trained alumen ustum image segmentation model for image segmentation, and acquiring alumen ustum contour information of the target object;
and acquiring the position information of the alum flocs of the target object according to the alum floc outline information.
2. The method for processing an alumen ustum image according to claim 1, wherein the alumen ustum contour information includes an alumen ustum contour region, a region area, an alumen ustum image gray variance, an alumen ustum image gradient, an alumen ustum image kurtosis, an alumen ustum image entropy, and an alumen ustum image fractal dimension.
3. The method for processing alum blossom image according to claim 1, wherein underwater alum blossom image sample data is obtained and subjected to data labeling to form an alum blossom image data set, and the alum blossom image data set is subjected to image preprocessing, wherein the preprocessing comprises image defogging processing, and the image defogging processing comprises obtaining a dark channel image by obtaining a minimum value in RGB components of each pixel in the image.
4. The alum blossom image processing method according to claim 3, wherein the dark channel image is obtained by the following formula:
wherein, JdarkFor each channel of the dark channel image; j. the design is a squarecFor each channel of the color image; Ω (x) is a window centered on pixel x.
5. The method for processing alumen ustum image of claim 3, wherein the preprocessing further comprises a low-light image enhancement process comprising
Decomposing the image into an incident image and a reflected image;
and removing the incident attribute in the image and keeping the original image attribute.
6. The method for processing an image of alum blossom according to claim 2, wherein a bounding rectangle frame of the outline of the target object is obtained from the alum blossom outline information, and alum blossom position information is obtained from the coordinates of the bounding rectangle frame.
7. The method of claim 6, wherein an alumen ustum outline region and a region area of the target object are determined from the bounding rectangle, a water background region is determined from the alumen ustum outline region, and the alumen ustum and the water background in the target object image are subjected to image segmentation.
8. The method for processing alum blossom image as claimed in claim 2, wherein the alum blossom image segmentation model comprises
A contraction path for obtaining context information in the target object;
and the extended path is used for positioning the part needing to be segmented in the target object.
9. The method of claim 8, wherein a maximum pooling layer is connected after every two convolutional layers in the contraction path, and each convolutional layer is followed by down-sampling the original image of the target object by an activation function, and the number of channels is doubled at each down-sampling.
10. The method as claimed in claim 8, wherein the extended path includes a plurality of convolution layers for performing upsampling at each step, and each step of upsampling is added to the feature map corresponding to the contracted path, respectively, and the last layer of the alumen ustum image segmentation model is the convolution layer for converting the feature vectors of the channels into the number of required classification results.
11. The method for processing the alum blossom image as claimed in claim 2, wherein the pixel level features in the image are expanded, and the image segmentation is performed by combining local and global clues, wherein the expansion comprises fusing a plurality of different pyramid scale features, using a convolution level-reducing channel for each level, obtaining the size of the image before pooling through bilinear difference values, and finally merging the size of the image together.
12. An alum blossom image processing system, comprising:
the data sample module is used for acquiring underwater alum blossom image sample data and carrying out data annotation;
the image processing model is used for establishing an alum blossom image segmentation model according to the marked data and training;
the image acquisition module is used for acquiring a real-time image of the target object and inputting the real-time image into the trained alum blossom image segmentation model for image segmentation;
the alumen ustum image segmentation model acquires alumen ustum contour information of a target object, and acquires alumen ustum position information of the target object according to the alumen ustum contour information.
13. The alumen ustum image processing system of claim 12, wherein the alumen ustum image segmentation model comprises mutually symmetric:
a contraction path for obtaining context information in the target object;
and the extended path is used for positioning the part needing to be segmented in the target object.
14. The alumen ustum image processing system of claim 12, wherein a maximum pooling layer is connected after every two convolutional layers in the contraction path, and each convolutional layer is followed by down-sampling an original image of a target object by an activation function, and the number of channels is doubled at each down-sampling.
15. The system of claim 12, wherein the extension path comprises convolution layers for performing upsampling at each step, and each step of upsampling is added to the feature map corresponding to the contraction path.
16. The system of claim 12, wherein the fir image segmentation model generates a predictive segmentation result of the verification picture after each batch of training is completed, trains model parameters according to the annotation data and the predictive segmentation result, and determines the final recognition accuracy.
17. The alumen ustum image processing system according to claim 16, wherein a model with the highest accuracy is selected as the optimal model based on an image segmentation result of the alumen ustum image segmentation model.
18. A computer-readable storage medium having stored thereon a computer program, characterized in that: the program when executed by a processor implements the method of any one of claims 1 to 9.
19. An electronic terminal, comprising: a processor and a memory;
the memory is for storing a computer program and the processor is for executing the computer program stored by the memory to cause the terminal to perform the method of any of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010703199.XA CN111833369A (en) | 2020-07-21 | 2020-07-21 | Alum image processing method, system, medium and electronic device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010703199.XA CN111833369A (en) | 2020-07-21 | 2020-07-21 | Alum image processing method, system, medium and electronic device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111833369A true CN111833369A (en) | 2020-10-27 |
Family
ID=72923759
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010703199.XA Pending CN111833369A (en) | 2020-07-21 | 2020-07-21 | Alum image processing method, system, medium and electronic device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111833369A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112808603A (en) * | 2020-12-22 | 2021-05-18 | 南京林业大学 | Fresh cut flower sorting device and method based on RealSense camera |
| CN113429013A (en) * | 2021-06-03 | 2021-09-24 | 阿里巴巴新加坡控股有限公司 | Method for determining coagulant addition amount and method for determining compound addition amount |
| CN113688751A (en) * | 2021-08-30 | 2021-11-23 | 上海城投水务（集团）有限公司制水分公司 | Method and device for analyzing alum blossom characteristics by using image recognition technology |
| CN115147617A (en) * | 2022-09-06 | 2022-10-04 | 聊城集众环保科技有限公司 | Intelligent sewage treatment monitoring method based on computer vision |
| CN117237356A (en) * | 2023-11-15 | 2023-12-15 | 山东泗水丰田农药有限公司 | Wastewater treatment control method for pesticide production |
| CN117929375A (en) * | 2024-03-21 | 2024-04-26 | 武汉奥恒胜科技有限公司 | Water quality detection method and water quality detector based on image processing |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101783012A (en) * | 2010-04-06 | 2010-07-21 | 中南大学 | Automatic image defogging method based on dark primary colour |
| CN109242869A (en) * | 2018-09-21 | 2019-01-18 | 科大讯飞股份有限公司 | A kind of image instance dividing method, device, equipment and storage medium |
| CN110766640A (en) * | 2019-11-05 | 2020-02-07 | 中山大学 | Image defogging method based on depth semantic segmentation |
| CN111294516A (en) * | 2020-02-27 | 2020-06-16 | 中冶赛迪重庆信息技术有限公司 | Alum image processing method and system, electronic device and medium |
-
2020
- 2020-07-21 CN CN202010703199.XA patent/CN111833369A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101783012A (en) * | 2010-04-06 | 2010-07-21 | 中南大学 | Automatic image defogging method based on dark primary colour |
| CN109242869A (en) * | 2018-09-21 | 2019-01-18 | 科大讯飞股份有限公司 | A kind of image instance dividing method, device, equipment and storage medium |
| CN110766640A (en) * | 2019-11-05 | 2020-02-07 | 中山大学 | Image defogging method based on depth semantic segmentation |
| CN111294516A (en) * | 2020-02-27 | 2020-06-16 | 中冶赛迪重庆信息技术有限公司 | Alum image processing method and system, electronic device and medium |
Non-Patent Citations (4)
| Title |
|---|
| HENGSHUANG ZHAO 等: "Pyramid Scene Parsing Network", 《2017 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR)》 * |
| 周宇亮: "基于深度学习的磁共振吉布斯伪影移除和并行成像", 《中国优秀博硕士学位论文全文数据库(硕士) 工程科技Ⅱ辑》 * |
| 李嘉冀: "基于模糊逻辑注意力机制U-Net乳腺肿瘤图像分割", 《中国优秀博硕士学位论文全文数据库(硕士) 工程科技Ⅱ辑》 * |
| 王一涵: "雾天图像增强方法研究", 《中国优秀博硕士学位论文全文数据库(硕士) 信息科技辑》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112808603A (en) * | 2020-12-22 | 2021-05-18 | 南京林业大学 | Fresh cut flower sorting device and method based on RealSense camera |
| CN112808603B (en) * | 2020-12-22 | 2021-10-12 | 南京林业大学 | Fresh cut flower sorting device and method based on RealSense camera |
| CN113429013A (en) * | 2021-06-03 | 2021-09-24 | 阿里巴巴新加坡控股有限公司 | Method for determining coagulant addition amount and method for determining compound addition amount |
| CN113429013B (en) * | 2021-06-03 | 2022-12-27 | 阿里巴巴新加坡控股有限公司 | Method for determining coagulant addition amount and method for determining compound addition amount |
| CN113688751A (en) * | 2021-08-30 | 2021-11-23 | 上海城投水务（集团）有限公司制水分公司 | Method and device for analyzing alum blossom characteristics by using image recognition technology |
| WO2023029117A1 (en) * | 2021-08-30 | 2023-03-09 | 上海城市水资源开发利用国家工程中心有限公司 | Method and apparatus for analyzing alum floc feature by using image recognition technology |
| CN115147617A (en) * | 2022-09-06 | 2022-10-04 | 聊城集众环保科技有限公司 | Intelligent sewage treatment monitoring method based on computer vision |
| CN117237356A (en) * | 2023-11-15 | 2023-12-15 | 山东泗水丰田农药有限公司 | Wastewater treatment control method for pesticide production |
| CN117237356B (en) * | 2023-11-15 | 2024-02-23 | 山东泗水丰田农药有限公司 | Wastewater treatment control method for pesticide production |
| CN117929375A (en) * | 2024-03-21 | 2024-04-26 | 武汉奥恒胜科技有限公司 | Water quality detection method and water quality detector based on image processing |
| CN117929375B (en) * | 2024-03-21 | 2024-06-04 | 武汉奥恒胜科技有限公司 | Water quality detection method and water quality detector based on image processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111833369A (en) | Alum image processing method, system, medium and electronic device | |
| CN110163080B (en) | Face key point detection method and device, storage medium and electronic equipment | |
| Aquino et al. | Grapevine flower estimation by applying artificial vision techniques on images with uncontrolled scene and multi-model analysis | |
| CN111046880A (en) | Infrared target image segmentation method and system, electronic device and storage medium | |
| CN113498528A (en) | Image defect determining method and device, electronic equipment and storage medium | |
| CN112183482A (en) | Dangerous driving behavior recognition method, device and system and readable storage medium | |
| CN110766007B (en) | Certificate shielding detection method, device, equipment and readable storage medium | |
| KR102674065B1 (en) | Defect detection system and method through image completion based on artificial intelligence-based denoising | |
| CN112417955A (en) | Patrol video stream processing method and device | |
| CN107563290A (en) | A kind of pedestrian detection method and device based on image | |
| CN113537037A (en) | Pavement disease identification method, system, electronic device and storage medium | |
| CN116740728B (en) | Dynamic acquisition method and system for wafer code reader | |
| CN115439804A (en) | Monitoring method and device for high-speed rail maintenance | |
| CN115239581A (en) | Image processing method and related device | |
| CN116934762B (en) | System and method for detecting surface defects of lithium battery pole piece | |
| CN111179289B (en) | Image segmentation method suitable for webpage length graph and width graph | |
| CN116977840A (en) | Marine organism target detection method, system, storage medium and equipment | |
| CN116311212A (en) | Ship number identification method and device based on high-speed camera and in motion state | |
| CN116228769A (en) | Device and method suitable for flaw detection of steel wire braided tube | |
| CN116152191A (en) | Display screen crack defect detection method, device and equipment based on deep learning | |
| CN115358981A (en) | Glue defect determining method, device, equipment and storage medium | |
| CN115661803A (en) | Image definition detection method, electronic device, and computer-readable storage medium | |
| CN114781489A (en) | Semantic segmentation model training method, semantic segmentation method and electronic equipment | |
| CN111815658B (en) | Image recognition method and device | |
| CN114004809A (en) | Skin image processing method, device, electronic equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 401329 No. 5-6, building 2, No. 66, Nongke Avenue, Baishiyi Town, Jiulongpo District, Chongqing Applicant after: MCC CCID information technology (Chongqing) Co.,Ltd. Address before: 20-24 / F, No.7 Longjing Road, North New District, Yubei District, Chongqing Applicant before: CISDI CHONGQING INFORMATION TECHNOLOGY Co.,Ltd. |
|
| CB02 | Change of applicant information | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201027 |
|
| RJ01 | Rejection of invention patent application after publication |