CN116824258A - Construction site smoke dust detection method based on back projection - Google Patents
Construction site smoke dust detection method based on back projection Download PDFInfo
- Publication number
- CN116824258A CN116824258A CN202310792252.1A CN202310792252A CN116824258A CN 116824258 A CN116824258 A CN 116824258A CN 202310792252 A CN202310792252 A CN 202310792252A CN 116824258 A CN116824258 A CN 116824258A
- Authority
- CN
- China
- Prior art keywords
- sample image
- image
- histogram
- distribution
- mask
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 89
- 239000000779 smoke Substances 0.000 title claims abstract description 48
- 239000000428 dust Substances 0.000 title claims abstract description 33
- 238000010276 construction Methods 0.000 title claims abstract description 19
- 239000011159 matrix material Substances 0.000 claims abstract description 8
- 238000000034 method Methods 0.000 claims description 14
- 238000001914 filtration Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000012549 training Methods 0.000 abstract description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000009792 diffusion process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 238000013100 final test Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/42—Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
A construction site smoke dust detection method based on back projection relates to the field of environment detection. The invention aims to solve the problems that the existing smoke detection method is short in detection distance, the model precision depends on the training sample amount, the smoke detection sample is difficult to obtain and the detection accuracy is low. The invention comprises the following steps: acquiring a distribution probability form histogram of a sample image; learning the distribution probability of the sample image through the combination and difference operation; acquiring an image to be detected, back projecting the image to be detected by utilizing the learned distribution probability histogram, acquiring a target probability distribution matrix, convolving the target probability distribution matrix by utilizing a Gaussian filter operator, and extracting regional target distribution probability information; comparing the regional target distribution probability information with a preset threshold value to obtain a preliminary identification result; and carrying out region growth on the mask in the initial identification result through the sliding window to obtain a mask detection result, thereby obtaining the smoke dust proportion in the image to be detected. The invention is used for detecting the smoke dust ratio of the construction site.
Description
Technical Field
The invention relates to the field of environment detection, in particular to a construction site smoke dust detection method based on back projection.
Background
The traditional dust recognition method mostly adopts physical signals for monitoring, such as detecting the concentration of PM2.5 and PM10 in the air through an air quality sensor to reflect the dust content in the air, but the recognition distance of the method is limited, the method is dependent on environment and is easy to be interfered by the environment, so that the smoke detection distance is short and the accuracy is low. At present, means such as a neural network is adopted to realize the learning and the identification of dust images, but the method is more dependent on a data set, however, the dust area is difficult to divide due to the fact that a training set is difficult to acquire, and the problem of low dust detection accuracy is caused.
Disclosure of Invention
The invention aims to solve the problems of short detection distance and low detection accuracy of the existing smoke detection method, and provides a construction site smoke detection method based on back projection.
The construction site smoke dust detection method based on back projection comprises the following specific processes:
step one, acquiring a sample image, thereby acquiring a distribution probability form histogram of the sample image;
step two, learning the distribution probability form histogram of the sample image obtained in the step one through cross-merging operation and difference operation to obtain a learned distribution probability form histogram;
thirdly, acquiring an image to be detected, performing back projection on the image to be detected by utilizing the two-dimensional distribution histogram in the learned distribution probability form acquired in the second step to acquire a target probability distribution matrix P, and then convolving the target probability distribution matrix P by utilizing a Gaussian filter operator to extract regional target distribution probability information;
comparing the regional target distribution probability information obtained in the step three with a preset threshold value, setting an image mask of a region with probability larger than the preset threshold value in the image to be detected to be 1, setting an image mask of a region with probability smaller than or equal to the preset threshold value in the image to be detected to be 0, and obtaining a preliminary recognition result; then, carrying out region growth on a mask in the initial identification result through a sliding window to obtain a mask detection result;
and step five, obtaining the smoke dust ratio in the image to be detected by using the mask detection result obtained in the step four.
Further, the step one of acquiring a sample image, thereby obtaining a distribution probability form histogram of the sample image, specifically:
step one, acquiring a sample image, converting the sample image into an HSV data format, and then extracting a hue channel H and a saturation channel S of the sample image;
the sample image includes: sample image H of positive example t Counterexample sample image H f ；
The positive sample image is a sample image containing a detection object;
the counterexample sample image is a sample image of a detection environment;
step one, scaling the range of the hue channel H of the sample image extracted in the step one by one to 0-360, and taking a as H channel unit interval and taking the x axis as the x axis; scaling the range of the saturation channel S to 0-255, taking b as the S channel interval, and taking the b as the y axis, thereby obtaining a HS channel two-dimensional distribution histogram of the sample image;
and step one, converting the HS channel two-dimensional distribution histogram obtained in the step one into a probability distribution form to obtain a distribution probability form histogram of the sample image.
Further, the distribution probability of each position in the distribution probability form histogram of the sample image is obtained by the following formula:
where H1 is the HS channel two-dimensional distribution histogram, (x, y) is a point coordinate in the two-dimensional distribution histogram, max (H1) is a maximum value in the HS channel two-dimensional distribution histogram, H' (x, y) is a distribution probability at (x, y), min (H1) is a minimum value in the HS channel two-dimensional distribution histogram, and H1 (x, y) is the HS channel two-dimensional distribution histogram at (x, y).
Further, in the second step, the distribution probability form histogram of the sample image obtained in the first step is learned by means of cross-combining operation and difference operation, so as to obtain a learned distribution probability form histogram, which is specifically as follows:
step two, carrying out intersection operation on the distribution probability form histogram of the sample image of the positive example, wherein the formula is as follows:
where n is the total number of blocks in the distribution probability form histogram of the sample image, j is the block index, H t1 (j) Is the t 1 The j-th block, H in the distribution probability form histogram of the sample image of the positive example t2 (j) Is the t 2 The j-th block in the distribution probability form histogram of the positive sample image;
step two, carrying out union operation on the distribution probability form histogram of the sample image of the positive example, wherein the formula is as follows:
step two, carrying out difference operation on the distribution probability form histogram of the positive example sample image and the distribution probability form histogram of the negative example sample image, wherein the difference operation is as follows:
further, the gaussian filter operator in the third step is as follows:
wherein sigma 2 Is the filtering variance, x 'is the transverse axis dimension of the filter, and y' is the longitudinal axis dimension of the filter.
Further, in the fourth step, the mask in the initial recognition result is subjected to region growth through a sliding window, so as to obtain a mask detection result, wherein the mask detection result is represented by the following formula:
wherein M (x 1, y 1) passes through the image mask to be detected of threshold screening,for M, model is the mask detection result, beta is the hyper-parameter, the abscissa of the pixels in the x1, y1 sliding window, ave is the mean, and std is the variance.
Further, the method comprises the steps of,
where Num is the number of marked pixels in the mask M and W is the sliding window length.
Further, the method comprises the steps of,
further, in the fifth step, the ratio of the smoke dust in the image to be detected is obtained by using the mask detection result obtained in the fourth step, specifically:
fifthly, corroding and expanding the mask detection result obtained in the fourth step, removing detection noise points, and obtaining a mask detection result Model 'after the detection noise points are removed, wherein the mask detection result Model' specifically comprises the following steps:
adopting linear spherical operators with the radius of 4, respectively adopting one-time opening operation and one-time closing operation for the mask detection result obtained in the step four, filling the holes of the mask of the image, and simultaneously removing discrete points to obtain a Model';
and fifthly, obtaining the smoke dust ratio in the image to be detected by using a mask detection result Model' after the detection noise points are removed.
Further, in the fifth step, the smoke and dust ratio in the image to be detected is obtained by using the mask detection result Model' after the detection noise is removed, and the following formula is adopted:
in the formula, row is the number of rows of the image to be measured, col is the number of columns of the image to be measured, and Sum () is the summation function.
The beneficial effects of the invention are as follows:
according to the invention, the recognition image is converted into a probability form of target distribution through a histogram back projection method, the probability of existence of smoke dust is reflected through convolution extraction of local distribution probability, the detection coverage area of the smoke dust is expanded through a growth algorithm, holes and noise points are eliminated through corrosion expansion, and finally the smoke dust-viewing coverage mask is obtained. The positive sample image of the invention selects a similar smoke sample image according to the characteristics of detected smoke; the counterexample template collects the environment image without smoke dust from the detection site without training by using a large number of data samples. According to the invention, the smoke and dust ratio is obtained through the segmentation and quantitative calculation of the smoke and dust, so that the smoke and dust ratio quantifies the diffusion degree of the smoke and dust. The invention has long detection distance and does not depend on the data set to divide the dust area, thereby improving the accuracy of smoke detection.
Drawings
FIG. 1 is a view of a test case image of the present invention;
FIG. 2 is a positive example template selected for use in the detection process of the present invention;
FIG. 3 is a two-dimensional histogram calculated after extracting HS channels from a template map;
FIG. 4 is a probability image of the target distribution of the detected image by back projection of the template histogram;
FIG. 5 is a range image of the original range of the mask, the range after growth, and the range after etching and expansion operations;
FIG. 6 is a final test result image;
fig. 7 is a flow chart of the present invention.
Detailed Description
The first embodiment is as follows: as shown in fig. 7, the specific process of the construction site smoke detection method based on back projection in this embodiment is:
step one, acquiring a sample image, thereby acquiring a distribution probability form histogram of the sample image:
step one, acquiring a sample image according to the detected environmental condition and the detected smoke dust type, converting the sample image into an HSV data format, and then extracting a hue channel H and a saturation channel S of the sample image;
the sample image includes: sample image H of positive example t Counterexample sample image H f ；
The positive sample image is a sample image containing a detection object, as shown in fig. 2;
the counterexample sample image is a sample image of a detection environment;
step one, scaling the range of the hue channel H of the sample image extracted in the step one by one to 0-360, and taking 2 as H channel unit intervals as an x axis; scaling the range of the saturation channel S to 0-255, taking 1 as the S channel interval, and taking the S channel interval as the y axis, so as to obtain a HS channel two-dimensional distribution histogram of the sample image, as shown in FIG. 3;
step one, converting the HS channel two-dimensional distribution histogram obtained in the step one into a probability distribution form to obtain a distribution probability form histogram of the sample image, as shown in fig. 4;
the probability of distribution for each location is obtained by the following formula:
where H1 is the HS channel two-dimensional distribution histogram, (x, y) is a point coordinate in the two-dimensional distribution histogram, max (H1) is a maximum value in the HS channel two-dimensional distribution histogram, H' (x, y) is a distribution probability at (x, y), min (H1) is a minimum value in the HS channel two-dimensional distribution histogram, and H1 (x, y) is the HS channel two-dimensional distribution histogram at (x, y).
Step two, learning the distribution probability form histogram of the sample image obtained in the step one through cross operation and difference operation to obtain a learned distribution probability form histogram, wherein the distribution probability form histogram is specifically as follows:
step two, performing intersection operation on the distribution probability form histogram of the sample image of the positive example to extract common features among the samples of the positive example, as shown in a formula (2):
where j is the block index, n is the total number of histogram blocks, H t1 (j) Is the t 1 The j-th block, H in the distribution probability form histogram of the sample image of the positive example t2 (j) Is the t 2 The j-th block in the distribution probability form histogram of the positive sample image; the detection precision can be improved through intersection operation, and interference caused by similar colors in the environment is reduced.
Step two, carrying out union operation on the distribution probability form histogram of the sample image of the positive example to generalize the detection result and improve the sensitivity of the model to smoke detection, wherein the sensitivity is shown in a formula (3):
the union operation can enable the detection template to have the characteristics of a plurality of positive sample images at the same time, and the condition of missing detection is reduced.
Step two, carrying out difference operation on the distribution probability form histogram of the positive example sample image and the distribution probability form histogram of the negative example sample image, reducing the false recognition of the model to the environment, and improving the accuracy rate, wherein the difference operation is shown in a formula (4):
wherein H is t (j) Is the j-th block in the histogram of the distribution probability form of the sample image, H f (j) Is the j-th block in the distribution probability form histogram of the counterexample sample image.
The difference operation keeps different hue and saturation intervals in the template and the environment, and the same part as the environment is deleted, so that the interference of the environment in the detection process is avoided.
Step three, acquiring a construction site video for identification and detection through monitoring, aerial photography and other ways, and performing frame extraction and sampling on the video according to identification requirements to obtain an image to be detected, as shown in fig. 1; performing back projection on the image to be detected by using the two-dimensional distribution histogram of the learned distribution probability form obtained in the second step to obtain a target probability distribution matrix P, and then convolving the target probability distribution matrix by using a Gaussian filter operator to extract regional target distribution probability information;
the gaussian filter operator has the following formula:
wherein the filtering variance sigma 2 Taking 1 and the radius of the filter taking 3. The radius of the filter can be adjusted according to the resolution and the requirement of the image, the larger the radius of the filter is, the stronger the noise filtering capability is, the higher the accuracy is, the detection range is reduced, x 'is the transverse axis size of the filter, and y' is the longitudinal axis size of the filter.
Comparing the regional target distribution probability information obtained in the step three with a preset threshold value, setting an image mask with probability larger than the preset threshold value region in the image to be detected to be 1, and keeping the image mask with probability smaller than or equal to the preset threshold value region to be 0 value, so as to obtain a preliminary recognition result; then, carrying out region growth on a mask in the initial identification result through a sliding window to obtain a mask detection result, as shown in fig. 5-6;
the sliding window takes a 21 multiplied by 21 rectangular window, the average value and the variance of the probability of the recognition target in the window are calculated after the image is expanded, and the mask growth is carried out on the rest pixels contained in the range through the following formulas (6), (7) and (8):
wherein P is the target distribution probability obtained by back projection, M (x 1, y 1) passes through the image mask to be detected of threshold screening,m is not, namely, the undetected image mask to be detected, W is the current pixel interval of the sliding window, and Num is the number of marked pixels in the mask M; model is the final detected mask result, beta is the hyper-parameter, the abscissa of the pixels in the x1, y1 sliding window, ave is the mean, and std is the variance.
Step five, obtaining the smoke dust ratio in the image to be detected by using the mask detection result obtained in the step four, wherein the specific steps are as follows:
fifthly, corroding and expanding the mask detection result obtained in the fourth step, removing detection noise points, and obtaining a mask detection result Model 'after the detection noise points are removed, wherein the mask detection result Model' specifically comprises the following steps:
the erosion and expansion operator adopts a linear spherical operator with the radius of 4, and adopts one-time open operation and one-time close operation respectively, so that the holes of the image mask are filled, and discrete points are removed;
step five, calculating the smoke dust ratio C of the image to be detected, wherein the calculation formula is as follows (9):
in the formula, row is the number of lines of the image to be detected, col is the number of columns of the image to be detected, model' is the mask detection result after detection noise is removed, and Sum () is the summation function.
And judging whether dust emission, fire and other conditions exist in the current image by setting a smoke early warning threshold value, and obtaining the smoke ratio condition.
Claims (10)
1. The construction site smoke dust detection method based on back projection is characterized by comprising the following specific processes of:
step one, acquiring a sample image, thereby acquiring a distribution probability form histogram of the sample image;
step two, learning the distribution probability form histogram of the sample image obtained in the step one through cross-merging operation and difference operation to obtain a learned distribution probability form histogram;
thirdly, acquiring an image to be detected, performing back projection on the image to be detected by utilizing the two-dimensional distribution histogram in the learned distribution probability form acquired in the second step to acquire a target probability distribution matrix R, and then convolving a target probability distribution P matrix by utilizing a Gaussian filter operator to extract regional target distribution probability information;
comparing the regional target distribution probability information obtained in the step three with a preset threshold value, setting an image mask with probability larger than a preset threshold value region in the image to be detected to 1, setting an image mask with probability smaller than or equal to the preset threshold value region in the image to be detected to 0, and obtaining a preliminary recognition result; then, carrying out region growth on a mask in the initial identification result through a sliding window to obtain a mask detection result;
and step five, obtaining the smoke dust ratio in the image to be detected by using the mask detection result obtained in the step four.
2. The back projection-based construction site smoke detection method as defined in claim 1, wherein: the step one of obtaining a sample image, thereby obtaining a distribution probability form histogram of the sample image, specifically:
step one, acquiring a sample image, converting the sample image into an HSV data format, and then extracting a hue channel H and a saturation channel S of the sample image;
the sample image includes: sample image H of positive example t Counterexample sample image H f ；
The positive sample image is a sample image containing a detection object;
the counterexample sample image is a sample image of a detection environment;
step one, scaling the range of the hue channel H of the sample image extracted in the step one by one to 0-360, and taking a as H channel unit interval and taking the x axis as the x axis; scaling the range of the saturation channel S to 0-255, taking b as the S channel interval, and taking the b as the y axis, thereby obtaining a HS channel two-dimensional distribution histogram of the sample image;
and step one, converting the HS channel two-dimensional distribution histogram obtained in the step one into a probability distribution form to obtain a distribution probability form histogram of the sample image.
3. The back projection-based construction site smoke detection method as set forth in claim 2, wherein: the distribution probability of each position in the distribution probability form histogram of the sample image is obtained by the following formula:
wherein H1 is a two-dimensional distribution histogram of the HS channel, (x, y) is a point coordinate in the two-dimensional distribution histogram, max (H1) is a maximum value in the two-dimensional distribution histogram of the HS channel, H ′ (x, y) is the probability of distribution at (x, y), min (H1) is the minimum in the two-dimensional distribution histogram of HS channels, and H1 (x, y) is the two-dimensional distribution histogram of HS channels at (x, y).
4. A method of back projection-based construction site smoke detection as defined in claim 3, wherein: and in the second step, the distribution probability form histogram of the sample image obtained in the first step is learned through the cross-merging operation and the difference operation, so that the learned distribution probability form histogram is obtained, and the method specifically comprises the following steps:
step two, carrying out intersection operation on the distribution probability form histogram of the sample image of the positive example, wherein the formula is as follows:
where n is the total number of blocks in the distribution probability form histogram of the sample image, j is the block index, H t1 (j) Is the t 1 The j-th block, H in the distribution probability form histogram of the sample image of the positive example t2 (j) Is the t 2 The j-th block in the distribution probability form histogram of the positive sample image;
step two, carrying out union operation on the distribution probability form histogram of the sample image of the positive example, wherein the formula is as follows:
step two, carrying out difference operation on the distribution probability form histogram of the positive example sample image and the distribution probability form histogram of the negative example sample image, wherein the difference operation is as follows:
wherein H is t (j) Is the j-th block in the histogram of the distribution probability form of the sample image, H f (j) Is the j-th block in the distribution probability form histogram of the counterexample sample image.
5. The back projection-based construction site smoke detection method as defined in claim 4, wherein: and (3) the Gaussian filter operator in the step three is as follows:
wherein sigma 2 Is the filtering variance, x 'is the transverse axis dimension of the filter, and y' is the longitudinal axis dimension of the filter.
6. The back projection-based construction site smoke detection method according to claim 5, wherein: and step four, performing region growth on a mask in the initial recognition result through a sliding window to obtain a mask detection result, wherein the mask detection result is represented by the following formula:
wherein M (x 1, y 1) passes through the image mask to be detected of threshold screening,for M, model is the mask detection result, beta is the hyper-parameter, the abscissa of the pixels in the x1, y1 sliding window, ave is the mean, and std is the variance.
7. The back projection-based construction site smoke detection method as defined in claim 6, wherein:
where Num is the number of marked pixels in the mask M and W is the sliding window length.
8. The back projection-based construction site smoke detection method as defined in claim 7, wherein:
9. the back projection-based construction site smoke detection method as defined in claim 8, wherein: in the fifth step, the ratio of smoke dust in the image to be detected is obtained by using the mask detection result obtained in the fourth step, specifically:
fifthly, corroding and expanding the mask detection result obtained in the fourth step, removing detection noise points, and obtaining a mask detection result Model 'after the detection noise points are removed, wherein the mask detection result Model' specifically comprises the following steps:
adopting linear spherical operators with the radius of 4, respectively adopting one-time opening operation and one-time closing operation for the mask detection result obtained in the step four, filling the holes of the mask of the image, and simultaneously removing discrete points to obtain a Model';
and fifthly, obtaining the smoke dust ratio in the image to be detected by using a mask detection result Model' after the detection noise points are removed.
10. The back projection-based construction site smoke detection method as set forth in claim 9, wherein: and in the fifth step, the mask detection result Model' after the detection noise point is removed is utilized to obtain the smoke dust ratio in the image to be detected, and the following formula is adopted:
in the formula, row is the number of rows of the image to be measured, col is the number of columns of the image to be measured, and Sum () is the summation function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310792252.1A CN116824258A (en) | 2023-06-30 | 2023-06-30 | Construction site smoke dust detection method based on back projection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310792252.1A CN116824258A (en) | 2023-06-30 | 2023-06-30 | Construction site smoke dust detection method based on back projection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116824258A true CN116824258A (en) | 2023-09-29 |
Family
ID=88121715
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310792252.1A Pending CN116824258A (en) | 2023-06-30 | 2023-06-30 | Construction site smoke dust detection method based on back projection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116824258A (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070083114A1 (en) * | 2005-08-26 | 2007-04-12 | The University Of Connecticut | Systems and methods for image resolution enhancement |
| CN104077779A (en) * | 2014-07-04 | 2014-10-01 | 中国航天科技集团公司第五研究院第五一三研究所 | Moving object statistical method with Gaussian background model and mean value shift tracking combined |
| CN105469105A (en) * | 2015-11-13 | 2016-04-06 | 燕山大学 | Cigarette smoke detection method based on video monitoring |
| US20170300778A1 (en) * | 2016-04-13 | 2017-10-19 | Sony Corporation | Object tracking device and method |
| CN108765443A (en) * | 2018-05-22 | 2018-11-06 | 杭州电子科技大学 | A kind of mark enhancing processing method of adaptive color Threshold segmentation |
| JP2019046039A (en) * | 2017-08-31 | 2019-03-22 | ホーチキ株式会社 | Fire detection device and fire detection method |
| JP2019178982A (en) * | 2018-03-30 | 2019-10-17 | 日本製鉄株式会社 | Dust fall-amount estimation method |
| CN110379514A (en) * | 2019-07-12 | 2019-10-25 | 天津市德安圣保安全卫生评价监测有限公司 | A kind of welding fume occupational disease hazards methods of risk assessment |
| CN110909756A (en) * | 2018-09-18 | 2020-03-24 | 苏宁 | Convolutional neural network model training method and device for medical image recognition |
| CN111210452A (en) * | 2019-12-30 | 2020-05-29 | 西南交通大学 | Certificate photo portrait segmentation method based on graph segmentation and mean shift |
| CN113033892A (en) * | 2021-03-23 | 2021-06-25 | 河海大学 | Dynamic evaluation method for credit of main body of construction market under government supervision view angle |
| CN113496159A (en) * | 2020-03-20 | 2021-10-12 | 昆明理工大学 | Multi-scale convolution and dynamic weight cost function smoke target segmentation method |
| WO2021254205A1 (en) * | 2020-06-17 | 2021-12-23 | 苏宁易购集团股份有限公司 | Target detection method and apparatus |
| CN115018751A (en) * | 2021-03-03 | 2022-09-06 | 上海新氦类脑智能科技有限公司 | Crack detection method and system based on Bayesian density analysis |
| CN115731188A (en) * | 2022-11-22 | 2023-03-03 | 浙江浙大鸣泉科技有限公司 | Method for detecting smoke in ambient air based on Ringelmann blackness |
-
2023
- 2023-06-30 CN CN202310792252.1A patent/CN116824258A/en active Pending
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070083114A1 (en) * | 2005-08-26 | 2007-04-12 | The University Of Connecticut | Systems and methods for image resolution enhancement |
| CN104077779A (en) * | 2014-07-04 | 2014-10-01 | 中国航天科技集团公司第五研究院第五一三研究所 | Moving object statistical method with Gaussian background model and mean value shift tracking combined |
| CN105469105A (en) * | 2015-11-13 | 2016-04-06 | 燕山大学 | Cigarette smoke detection method based on video monitoring |
| US20170300778A1 (en) * | 2016-04-13 | 2017-10-19 | Sony Corporation | Object tracking device and method |
| JP2019046039A (en) * | 2017-08-31 | 2019-03-22 | ホーチキ株式会社 | Fire detection device and fire detection method |
| JP2019178982A (en) * | 2018-03-30 | 2019-10-17 | 日本製鉄株式会社 | Dust fall-amount estimation method |
| CN108765443A (en) * | 2018-05-22 | 2018-11-06 | 杭州电子科技大学 | A kind of mark enhancing processing method of adaptive color Threshold segmentation |
| CN110909756A (en) * | 2018-09-18 | 2020-03-24 | 苏宁 | Convolutional neural network model training method and device for medical image recognition |
| CN110379514A (en) * | 2019-07-12 | 2019-10-25 | 天津市德安圣保安全卫生评价监测有限公司 | A kind of welding fume occupational disease hazards methods of risk assessment |
| CN111210452A (en) * | 2019-12-30 | 2020-05-29 | 西南交通大学 | Certificate photo portrait segmentation method based on graph segmentation and mean shift |
| CN113496159A (en) * | 2020-03-20 | 2021-10-12 | 昆明理工大学 | Multi-scale convolution and dynamic weight cost function smoke target segmentation method |
| WO2021254205A1 (en) * | 2020-06-17 | 2021-12-23 | 苏宁易购集团股份有限公司 | Target detection method and apparatus |
| CN115018751A (en) * | 2021-03-03 | 2022-09-06 | 上海新氦类脑智能科技有限公司 | Crack detection method and system based on Bayesian density analysis |
| CN113033892A (en) * | 2021-03-23 | 2021-06-25 | 河海大学 | Dynamic evaluation method for credit of main body of construction market under government supervision view angle |
| CN115731188A (en) * | 2022-11-22 | 2023-03-03 | 浙江浙大鸣泉科技有限公司 | Method for detecting smoke in ambient air based on Ringelmann blackness |
Non-Patent Citations (2)
| Title |
|---|
| SWEDER VAN WIJNBERGEN等: "Fiscal sustainability，volatility and oil wealth:A stochastic analysis of fiscal spending rules", 《TINBERGEN INSTITUTE DISCUSSION PAPER》, 30 April 2011 (2011-04-30), pages 1 - 31 * |
| 戴凌峰: "基于机器视觉的按压泵缺陷检测系统研究与实现", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》, no. 02, 15 February 2021 (2021-02-15), pages 029 - 497 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111310558B (en) | Intelligent pavement disease extraction method based on deep learning and image processing method | |
| CN107833220B (en) | Fabric defect detection method based on deep convolutional neural network and visual saliency | |
| CN112149543B (en) | Building dust recognition system and method based on computer vision | |
| CN109035273B (en) | Image signal fast segmentation method of immunochromatography test paper card | |
| CN112488046B (en) | Lane line extraction method based on high-resolution images of unmanned aerial vehicle | |
| CN110415208A (en) | A kind of adaptive targets detection method and its device, equipment, storage medium | |
| CN110175556B (en) | Remote sensing image cloud detection method based on Sobel operator | |
| CN111598098A (en) | Water gauge water line detection and effectiveness identification method based on full convolution neural network | |
| CN114841997B (en) | Method for detecting surface defects of sealing surface of valve base | |
| CN115731493A (en) | Rainfall micro physical characteristic parameter extraction and analysis method based on video image recognition | |
| CN111768455A (en) | Image-based wood region and dominant color extraction method | |
| CN111860531A (en) | Raise dust pollution identification method based on image processing | |
| CN114332781A (en) | Intelligent license plate recognition method and system based on deep learning | |
| CN116228756B (en) | Method and system for detecting bad points of camera in automatic driving | |
| CN111353968A (en) | Infrared image quality evaluation method based on blind pixel detection and analysis | |
| CN113065454A (en) | High-altitude parabolic target identification and comparison method and device | |
| CN111275687B (en) | Fine-grained image stitching detection method based on connected region marks | |
| CN115797299B (en) | Defect detection method of optical composite film | |
| CN116824258A (en) | Construction site smoke dust detection method based on back projection | |
| CN116563205A (en) | Wheat spike counting detection method based on small target detection and improved YOLOv5 | |
| CN114612333B (en) | Microscopic image-based micro-capsule algae cell statistical method | |
| CN106295642B (en) | License plate positioning method based on fault tolerance rate and texture features | |
| CN111523583B (en) | Method for automatically identifying and classifying equipment nameplate photos by using unmanned aerial vehicle | |
| CN115082444A (en) | Copper pipe weld defect detection method and system based on image processing | |
| CN111968136A (en) | Coal rock microscopic image analysis method and analysis system |
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 |