CN114332584A - Lake water surface floater identification method and medium based on image processing - Google Patents
Lake water surface floater identification method and medium based on image processing Download PDFInfo
- Publication number
- CN114332584A CN114332584A CN202111540531.6A CN202111540531A CN114332584A CN 114332584 A CN114332584 A CN 114332584A CN 202111540531 A CN202111540531 A CN 202111540531A CN 114332584 A CN114332584 A CN 114332584A
- Authority
- CN
- China
- Prior art keywords
- lake
- water surface
- image
- floaters
- floater
- 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
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 67
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000012545 processing Methods 0.000 title claims abstract description 32
- 208000034699 Vitreous floaters Diseases 0.000 claims abstract description 75
- 238000012549 training Methods 0.000 claims abstract description 20
- 238000013528 artificial neural network Methods 0.000 claims abstract description 17
- 238000000605 extraction Methods 0.000 claims abstract description 13
- 238000007781 pre-processing Methods 0.000 claims abstract description 10
- 238000012360 testing method Methods 0.000 claims description 14
- 238000005286 illumination Methods 0.000 claims description 13
- 238000001914 filtration Methods 0.000 claims description 10
- 238000012937 correction Methods 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 4
- 238000002372 labelling Methods 0.000 claims description 3
- 230000000877 morphologic effect Effects 0.000 claims description 3
- 238000003062 neural network model Methods 0.000 claims description 3
- 238000004088 simulation Methods 0.000 claims description 3
- 238000003709 image segmentation Methods 0.000 claims description 2
- 238000013135 deep learning Methods 0.000 abstract description 2
- 230000000694 effects Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000004140 cleaning Methods 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides a method and a medium for identifying a lake water surface floater based on image processing, wherein the method comprises the following steps: an image acquisition platform is built, different types of floaters in the water surface are acquired, and a camera is connected with a control computer; preprocessing the acquired original image and storing the preprocessed original image in a control computer; repeating the steps, collecting and measuring the floaters on the water surface of the lake for multiple times, and storing the obtained image data in a control computer; training image data, and establishing a deep neural network; and identifying the floaters on the water surface of the lake through a deep neural network to obtain the types of the floaters on the water surface of the lake. The invention trains the floater target feature extraction information through deep learning, further predicts the floater under the image target feature extraction information, and accurately identifies the floater on the water surface.
Description
Technical Field
The invention relates to the field of image recognition, in particular to a method and a medium for recognizing a lake water surface floater based on image processing.
Background
The technology for identifying the floaters on the water surface of the lake, which is used as a water surface floaters image identification technology for replacing human eyes, has been widely applied to a plurality of fields such as identification and classification of the floaters on the water surface of the lake, cleaning of the floaters on the water surface of the lake and the like, and is particularly prominent in the identification and cleaning of the floaters on the polluted water surface. However, due to the rapid development of machine vision and image recognition and processing technologies, the interior of the industry is lack of unified standards, and the actual performance of the identification method of the lake water surface floating objects is lack of an effective evaluation mode and detection means.
An important advantage of the technology for identifying the water surface floaters in the lake is that multiple targets are identified simultaneously, but the problem of how to improve the extraction of the target characteristic information of the water surface floaters is faced to realize the simultaneous identification of the multiple targets. An important index of the dynamic performance of the lake water surface floater identification method is target characteristic extraction information. In actual measurement, the dynamic performance of the method for identifying the floating objects on the lake surface is greatly influenced by the image target feature extraction information. If the target feature extraction information is not well extracted, phenomena such as missing identification or error identification can be generated, and the advantages of simultaneous multi-target identification do not exist any more. Therefore, the accuracy of the extraction information of the optimized target characteristics is improved, so that the identification of the lake water surface floaters is improved, and the method is of great importance to the development of the identification technology of the lake water surface floaters.
Disclosure of Invention
The purpose of the invention is as follows: the invention aims to provide a method and a medium for identifying the floaters on the water surface of a lake based on image processing, which can conveniently and accurately identify the floaters on the water surface of the lake.
The technical scheme is as follows: a lake water surface floater identification method based on image processing comprises the following steps:
step 1: an image acquisition platform is built, different types of floaters in the water surface are acquired, and a camera is connected with a control computer;
step 2: preprocessing the acquired original image and storing the preprocessed original image in a control computer;
and step 3: repeating the step 1 and the step 2, collecting and measuring the floaters on the water surface of the lake for multiple times, and storing the obtained image data in a control computer;
and 4, step 4: training image data, and establishing a deep neural network; and identifying the floaters on the water surface of the lake through a deep neural network to obtain the types of the floaters on the water surface of the lake.
Further, step 2 specifically includes:
step 2.1: performing feature extraction and image segmentation on the preprocessed image, highlighting the features of the floater through binary morphological operation, and performing feature extraction on the floater by utilizing a gray feature method;
step 2.2: and (5) repeating the step 2.1 to obtain original images, preprocessed images, characteristic images and segmentation images floating on the water surfaces of different types of lakes, and storing the processed images in the control computer.
Further, in step 2.1, the preprocessing includes graying processing, histogram equalization, filtering processing, binarization and background illumination non-uniformity correction.
Furthermore, a weighted average method is used for image graying, the weight of three channels is obtained through multiple simulation tests, and a formula for graying processing is as follows:
Gray＝0.07217B+0.71516G+0.21267R
where B denotes a blue channel, G denotes a green channel, and R denotes a red channel.
Further, histogram equalization is also called gray level equalization, and the gray level equalization formula is as follows:
wherein G isBDenotes the m equalized gray value, GARepresenting the gray value before equalization, GmaxRepresenting the maximum value of the gray before equalization, N the total number of pixels, FiThe number of pixels representing the i-th gray scale.
Further, the background illumination unevenness correction specifically adopts top-hat conversion to remove the background illumination.
Further, step 4 specifically includes:
step 4.1: labeling the lake water surface floaters obtained in the step 3 to obtain data sets of different types of lake water surface floaters, and dividing the data sets into a training set and a testing set;
step 4.2: establishing a deep neural network model, and training the lake floater image data obtained in the step (3) by using the deep neural network to construct an optimized deep neural network;
step 4.3: and detecting and identifying the water surface floater test set by using the trained deep neural network.
Further, in step 4.1, the ratio of training set to test set is 4: 1.
further, in step 1, during image acquisition, the acquired objects include plastic bottles, branches, and plastic bags.
In particular implementations, there are computer readable storage media comprising one or more programs for execution by one or more processors, the one or more programs including instructions for performing any of the methods described above.
Has the advantages that: compared with the prior art, the invention has the following remarkable advantages: the invention provides a lake water surface floater identification method based on image processing, wherein in the image preprocessing, a median filtering algorithm is used for reducing impact noise signals, and the filtering influence on step signals and harmonic signals in an image is small; aiming at the influence of illumination factors existing in the image background, the effect of removing background illumination by adopting top-hat transformation is the best; the floater target feature extraction information is trained through deep learning, the floater is predicted under the image target feature extraction information, the floater on the water surface is accurately identified, and then the floater on the water surface of the lake is conveniently cleaned.
Drawings
FIG. 1 is a flow chart of an image preprocessing section in the present invention;
FIG. 2 is a flow chart of an image recognition portion of the present invention;
FIG. 3 is a gray scale image of an original image according to the present invention;
FIG. 4 is a diagram of histogram equalization in the present invention;
FIG. 5 is a diagram of the equalized median filter of the present invention;
FIG. 6 is a graph of background illumination removal effect according to the present invention;
FIG. 7 is a binarized image after filtering in the present invention;
FIG. 8 is a segmentation of an image after feature selection in the present invention.
Detailed Description
The invention is further elucidated with reference to the drawings and the embodiments.
A lake water surface floater identification method based on image processing comprises the following steps:
step 1: and building a collection platform, wherein the camera is arranged on the camera support and is aligned to the lake surface with the water surface floating objects, and the camera is arranged and connected with the control computer. The camera is a high-definition camera supporting MJPEG format output and has adjustable focal length, and can provide 480P, 30FPS high-definition low-delay images;
step 2: an image acquisition step, namely selecting a proper lake water surface environment, sequentially putting three common water surface floaters, namely plastic bottles, branches and plastic bags, into the lake water surface, and acquiring images of the lake water surface floaters by using the camera of the camera by using the acquisition platform built in the step 1;
and step 3: an image processing step, namely, carrying out image processing on the original image acquired in the second step to finally obtain a processed image, and storing the processed image in a control computer;
step 3.1: and an image preprocessing step, namely preprocessing the water surface floater image obtained in the second step, wherein the main preprocessing comprises graying processing, histogram equalization, filtering processing, binarization and background illumination non-uniformity correction, and is shown in fig. 1. And respectively obtaining corresponding preprocessed images, and finally obtaining the best preprocessed image of the water surface floater. The image graying uses a weighted average method, namely, the weight of three channels is obtained through a plurality of times of simulation tests, and the graying processing effect is good. The grayscale image using the graying processing is shown in fig. 3, and the formula is as follows:
Gray＝0.07217B+0.71516G+0.21267R
and then histogram equalization is carried out on the grayed image, which is also called gray equalization. The equalized image is shown in fig. 4, and the gray level equalization formula is as follows:
wherein G isBRepresenting the equalized gray value, GARepresenting the gray value before equalization, GmaxRepresenting the maximum value of the gray before equalization, N the total number of pixels, FiThe number of pixels representing the i-th gray scale.
And performing median filtering on the image subjected to the histogram equalization processing, and simultaneously performing Gaussian and mean filtering on the image, and finally finding that the median filtering effect is the best, wherein the median filtering algorithm can reduce impact noise signals but has little influence on step signals and harmonic signals in the image, and the filtered image is as shown in FIG. 5.
For the influence of illumination factors existing in the image background, the effect of removing background illumination by adopting top-hat transformation is the best, and the definition is that the top-hat transformation of the gray-scale image is the opening operation of subtracting the image from the image, and the image after removing illumination is as shown in fig. 6.
And finally, carrying out binary operation on the processed image to obtain a binary image as shown in FIG. 7.
Step 3.2: a step of extracting the characteristics of the water surface floater, which is to extract the characteristics and segment the preprocessed image obtained in the step 3.1, highlight the characteristics of the floater by binary morphological operation and extract the characteristics of the floater by a gray characteristic method;
step 3.3: the image processing steps of different water surface floaters are repeated for different types of water surface floaters
And 3.2 and 3.3, obtaining original images, preprocessed images, characteristic images and segmentation images floating on the water surfaces of different types of lakes, and storing the processed images in a control computer, wherein the segmentation images of the images are shown in fig. 8.
And 4, step 4: and (3) image data step, repeating the step 2 and the step 3, collecting and measuring the floaters on the water surface of the lake for multiple times, sequentially putting three common floaters of a plastic bottle, a branch and a plastic bag on the water surface of the lake, and collecting the images of the three floaters on the water surface of the lake through the collection platform in the step 1. Respectively collecting 600 photos of the three water surface floaters through a camera, totaling 1800 photos, and storing the obtained image data in a control computer;
and 5: and an image identification step, namely training the image data obtained in the fourth step, establishing a deep neural network, and identifying the floaters on the water surface of the lake through the established deep neural network to obtain the types of the floaters on the water surface of the lake, wherein the flow of the method is shown in fig. 2.
Step 5.1: and 4, obtaining image data through the step 4, labeling the lake water surface floaters in the data set, obtaining data sets of three lake water surface floaters, and dividing the data sets into a training set and a testing set. Wherein the ratio of the training set to the test set is 4: 1, namely 1440 pictures exist in the training set, 360 pictures exist in the testing set, the training set is processed into 128 × 128 pictures in a unified mode, all the pictures are labeled, and due to the fact that the number of various types and the training difficulty of the data are different, for example, the recognition rate is low due to plastic bags, training needs to be conducted on the samples again after the training is finished;
step 5.2: training through the training set in the step 5.1 to establish a deep neural network model, and then training the lake floater image data obtained in the step 4 by using the deep neural network to establish an optimized deep neural network;
step 5.3: and (3) identifying the lake water surface floaters, namely using the lake water surface floaters test set of the image data obtained in the step 5.1, and detecting and identifying the water surface floaters test set by using the deep neural network trained in the step 5.2 to obtain the types of the lake water surface floaters.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The embodiments of the present invention are not described in detail, but are known in the art, and can be implemented by referring to the known techniques.
Claims (10)
1. A lake water surface floater identification method based on image processing is characterized by comprising the following steps:
step 1: an image acquisition platform is built, different types of floaters in the water surface are acquired, and a camera is connected with a control computer;
step 2: preprocessing the acquired original image and storing the preprocessed original image in a control computer;
and step 3: repeating the step 1 and the step 2, collecting and measuring the floaters on the water surface of the lake for multiple times, and storing the obtained image data in a control computer;
and 4, step 4: training image data, and establishing a deep neural network; and identifying the floaters on the water surface of the lake through a deep neural network to obtain the types of the floaters on the water surface of the lake.
2. The image processing-based lake surface floating object identification method according to claim 1, wherein the step 2 specifically comprises:
step 2.1: performing feature extraction and image segmentation on the preprocessed image, highlighting the features of the floater through binary morphological operation, and performing feature extraction on the floater by utilizing a gray feature method;
step 2.2: and (5) repeating the step 2.1 to obtain original images, preprocessed images, characteristic images and segmentation images floating on the water surfaces of different types of lakes, and storing the processed images in the control computer.
3. The method for identifying the floaters on the lake surface based on the image processing as claimed in claim 1, wherein in the step 2.1, the preprocessing comprises graying processing, histogram equalization, filtering processing, binarization and background illumination unevenness correction.
4. The image processing-based method for identifying the floaters on the lake surface as claimed in claim 3, wherein the graying of the image is performed by a weighted average method, the weight of three channels is obtained by a plurality of simulation tests, and the graying processing formula is as follows:
Gray＝0.07217B+0.71516G+0.21267R
where B denotes a blue channel, G denotes a green channel, and R denotes a red channel.
5. The image processing-based method for identifying the floaters on the lake surface as claimed in claim 3, wherein the histogram equalization is also called gray level equalization, and the gray level equalization formula is as follows:
wherein G isBRepresenting the equalized gray value, GARepresenting the gray value before equalization, GmaxRepresenting the maximum value of the gray before equalization, N the total number of pixels, FiThe number of pixels representing the i-th gray scale.
6. The image processing-based lake surface floating object identification method according to claim 3, wherein the background illumination unevenness correction specifically adopts top hat transformation to remove background illumination.
7. The image processing-based lake surface floating object identification method according to claim 1, wherein the step 4 specifically comprises:
step 4.1: labeling the lake water surface floaters obtained in the step 3 to obtain data sets of different types of lake water surface floaters, and dividing the data sets into a training set and a testing set;
step 4.2: establishing a deep neural network model, and training the lake floater image data obtained in the step (3) by using the deep neural network to construct an optimized deep neural network;
step 4.3: and detecting and identifying the water surface floater test set by using the trained deep neural network.
8. The image processing-based lake surface floating object identification method according to claim 7, wherein in the step 4.1, the ratio of the training set to the test set is 4: 1.
9. the image processing-based lake surface floating object identification method according to claim 1, wherein in the step 1, the collected objects include plastic bottles, branches and plastic bags during image collection.
10. A computer readable storage medium comprising one or more programs for execution by one or more processors, the one or more programs comprising instructions for performing the method of any of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111540531.6A CN114332584A (en) | 2021-12-16 | 2021-12-16 | Lake water surface floater identification method and medium based on image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111540531.6A CN114332584A (en) | 2021-12-16 | 2021-12-16 | Lake water surface floater identification method and medium based on image processing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114332584A true CN114332584A (en) | 2022-04-12 |
Family
ID=81052237
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111540531.6A Pending CN114332584A (en) | 2021-12-16 | 2021-12-16 | Lake water surface floater identification method and medium based on image processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114332584A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113435269A (en) * | 2021-06-10 | 2021-09-24 | 华东师范大学 | Improved water surface floating object detection and identification method and system based on YOLOv3 |
| CN113469097A (en) * | 2021-07-13 | 2021-10-01 | 大连理工大学人工智能大连研究院 | SSD (solid State disk) network-based real-time detection method for water surface floating object multiple cameras |
| US20210390728A1 (en) * | 2021-01-21 | 2021-12-16 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Object area measurement method, electronic device and storage medium |
-
2021
- 2021-12-16 CN CN202111540531.6A patent/CN114332584A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210390728A1 (en) * | 2021-01-21 | 2021-12-16 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Object area measurement method, electronic device and storage medium |
| CN113435269A (en) * | 2021-06-10 | 2021-09-24 | 华东师范大学 | Improved water surface floating object detection and identification method and system based on YOLOv3 |
| CN113469097A (en) * | 2021-07-13 | 2021-10-01 | 大连理工大学人工智能大连研究院 | SSD (solid State disk) network-based real-time detection method for water surface floating object multiple cameras |
Non-Patent Citations (1)
| Title |
|---|
| 许雪梅等: "《遥感数字图像处理教程 IDL编程实现》", 31 July 2021, 湖南科学技术出版社, pages: 394 - 399 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhao et al. | Cloud shape classification system based on multi-channel cnn and improved fdm | |
| CN108921130B (en) | Video key frame extraction method based on saliency region | |
| CN109978822B (en) | Banana maturity judging modeling method and judging method based on machine vision | |
| KR101403876B1 (en) | Method and Apparatus for Vehicle License Plate Recognition | |
| CN113592861B (en) | Bridge crack detection method based on dynamic threshold | |
| CN111626993A (en) | Image automatic detection counting method and system based on embedded FEFnet network | |
| CN114549981A (en) | Intelligent inspection pointer type instrument recognition and reading method based on deep learning | |
| CN110334760B (en) | Optical component damage detection method and system based on RESUnet | |
| CN113112483B (en) | Rigid contact net defect detection method and system based on similarity measurement | |
| CN113030244B (en) | Inversion imaging method and system for transmission line tower corrosion defect magnetic flux leakage detection signal | |
| CN108009567A (en) | A kind of automatic discriminating conduct of the fecal character of combination color of image and HOG and SVM | |
| CN111539330A (en) | Transformer substation digital display instrument identification method based on double-SVM multi-classifier | |
| CN112991271A (en) | Aluminum profile surface defect visual detection method based on improved yolov3 | |
| CN117152735A (en) | Tomato maturity grading method based on improved yolov5s | |
| CN110458019B (en) | Water surface target detection method for eliminating reflection interference under scarce cognitive sample condition | |
| Zhao et al. | Research on detection method for the leakage of underwater pipeline by YOLOv3 | |
| CN111950357A (en) | Marine water surface garbage rapid identification method based on multi-feature YOLOV3 | |
| CN110991374A (en) | Fingerprint singular point detection method based on RCNN | |
| CN110689003A (en) | Low-illumination imaging license plate recognition method and system, computer equipment and storage medium | |
| CN116704526B (en) | Staff scanning robot and method thereof | |
| CN113673396A (en) | Spore germination rate calculation method and device and storage medium | |
| CN115830514B (en) | Whole river reach surface flow velocity calculation method and system suitable for curved river channel | |
| CN114943869B (en) | Airport target detection method with enhanced style migration | |
| CN116503470A (en) | Method for segmenting, positioning and analyzing states of insulator string disc in infrared image | |
| CN114332584A (en) | Lake water surface floater identification method and medium based on image processing |
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 |