CN108734093B - Method for eliminating pseudo change of fixed-point land video monitoring - Google Patents
Method for eliminating pseudo change of fixed-point land video monitoring Download PDFInfo
- Publication number
- CN108734093B CN108734093B CN201810297035.4A CN201810297035A CN108734093B CN 108734093 B CN108734093 B CN 108734093B CN 201810297035 A CN201810297035 A CN 201810297035A CN 108734093 B CN108734093 B CN 108734093B
- Authority
- CN
- China
- Prior art keywords
- image
- area
- alarm
- snow
- value
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Closed-Circuit Television Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method for eliminating pseudo change of fixed-point land video monitoring, which comprises the following steps: establishing a local spring and autumn ploughing time and ploughing area information database; calculating the high brightness area ratio of the original background image; calculating the size of an alarm area range, and triggering a pseudo change detection mechanism if the area range is larger than a threshold value: whether the alarm is triggered by the false change is judged from three aspects of 'whether the alarm is rainy/snowy', 'whether the alarm is covered by snow' and 'whether the alarm is in agricultural production', and when at least one of the three judgments is 'true', the alarm is cancelled. The invention solves the problems of low efficiency and high cost of the traditional manual identification and operation; the detection accuracy is obviously improved; the interference caused by the ambient light and other environmental factors is reduced.
Description
Technical Field
The invention relates to a video signal processing method, in particular to a method for eliminating pseudo change of fixed-point land video monitoring.
Background
In order to effectively protect basic farmland and mineral resources and frighten illegal land behaviors, the illegal behaviors of the national resources are discovered, restrained and checked at the first time, and the fixed-point land video monitoring system is used for monitoring the behaviors. When the land is detected to be illegally used, the system generates alarm information and prevents illegal act of investigation and treatment. However, the following two problems exist in the actual monitoring process:
1. when snowing or heavy rain weather occurs, the monitored land area has snow cover or the monitoring picture is influenced by raindrops and snowing, which causes great interference to the land monitoring result and may have large-scale false alarm.
2. If the monitoring area is mainly the farmland, the condition that the land image characteristics are changed in a large area exists in spring ploughing and autumn harvesting seasons, the alarm can be triggered to generate false alarm.
Disclosure of Invention
The purpose of the invention is as follows: aiming at the problem that the existing land fixed-point monitoring is easily interfered by factors such as rain and snow weather or cultivation, the invention provides a method for eliminating the pseudo change of the fixed-point land video monitoring, which can obviously reduce the false alarm rate.
The technical scheme is as follows: a method for eliminating the pseudo-change of fixed-point land video monitoring comprises the following steps:
(1) storing the spring ploughing and autumn harvesting time and ploughing coverage area information of the monitoring area into a database;
(2) making a brightness histogram of an original image of the monitoring area, and calculating the range size of the high-brightness area;
(3) when the system gives an alarm, calculating the range of an alarm area;
(4) if the range of the alarm area exceeds 30% of the monitoring area, performing the step (5);
(5) and when at least one of the following three judgments is true, judging the alarm caused by the false change:
(5.1) judging whether snowfall/rain occurs: acquiring weather information from the network service, judging whether the weather information is a pseudo change caused by rain/snow by combining with the characteristics of rain/snow in the image, and if so, canceling the alarm;
(5.2) judging whether snow is accumulated: acquiring weather information in a week, judging whether a snowfall condition occurs in the week, extracting an area meeting the snow characteristics in an image, judging whether a monitored area is covered by snow, and if so, cancelling alarm;
(5.3) judging whether the activity is agricultural production activity: comparing the spring ploughing and autumn harvesting time in the database with the local real-time, judging whether the local is in agricultural production or not by combining the coincidence rate of the alarm area and the ploughing area, and if so, canceling the alarm;
and when the three judgments are all 'false', normally alarming.
The step (2) comprises the following steps:
(2.1) acquiring a screenshot of a monitoring picture under the condition of no rain or snow, and performing hsv channel separation on the acquired image;
(2.2) making a brightness histogram of the image to obtain the maximum brightness value P of the image;
(2.3) setting a threshold value T ═ a × P, wherein a is used for judging high brightness pixel points in the original image, a is more than or equal to 0.8 and less than or equal to 0.9, counting the number n of pixels with brightness larger than T, and n is the number of the high brightness pixel points in the original image.
The step (5.1) specifically comprises the following steps:
(5.1.1) graying the current monitoring picture by adopting a gray formula I ═ R + G + B)/3;
(5.1.2) performing median filtering on the grayscale image obtained in (5.1.1), and removing the characteristic similar to snowflakes or raindrops existing in the image;
(5.1.3) making a difference between the gray level images obtained in the step (5.1.1) and the gray level image obtained in the step (5.1.2), setting a threshold value T ' of a brightness difference value, comparing the difference with the threshold value T ' after the brightness values of two pixel points are made the difference, judging whether the pixel point meets the raindrop and snowflake characteristics, counting the number n ' of the pixels, and taking the threshold value T ' ' from 20 to 40;
(5.1.4) if the total number of pixels in the image is M x N, if
If the value of (1) is greater than the threshold value T '', judging that snow/rain exists in the image; the threshold value T '' is a threshold value of the ratio of the pixel point number meeting the snowflake feature to the total pixel point number of the whole image, if the ratio is larger than the threshold value T '', the characteristic pixels with a large number of suspected snowflakes exist in the image, and whether the image snows or not is judged; the threshold value T' takes a value of 3% -5%;
the step (5.2) specifically comprises the following steps:
(5.2.1) the system acquires local real-time weather information through the meteorological platform and judges whether snowfall weather occurs in one week;
(5.2.2) extracting a suspected snow area in the image according to the features of the snow; the component with the largest value in R, G, B three components of a pixel point is marked as Max, the minimum value is marked as Min, and when the (Max-Min) < alpha and (R + G + B)/3> beta are met, the pixel point can be judged to meet the color and brightness characteristics of the accumulated snow; the value of alpha is selected according to the contrast, and is generally selected to be less than 20; the value of beta is selected according to the maximum lightness P of the image, and is usually selected as beta-b-P, wherein b is used for judging high-lightness pixel points in the image, and b is more than or equal to 0.8 and less than or equal to 0.9;
(5.2.3) judging whether the coverage rate of the area meeting the snow characteristics to the picture is greater than s%, and is more than or equal to 20% and less than or equal to 50%, if so, indicating that the snow coverage condition exists in the monitored area at the moment, and stopping alarming; and selecting the specific value of s according to the snowfall condition of the local past year.
The step (5.3) specifically comprises the following steps:
(5.3.1) acquiring local real-time through network service, comparing the local real-time with the spring and autumn time in the database, and judging whether the local real-time is in the agricultural production period;
(5.3.2) comparing the alarm area with the cultivated land area stored in the step (1), if the coincidence rate of the alarm area and the cultivated land area exceeds c percent and is more than or equal to 70 percent and less than or equal to 95 percent, judging that the monitoring area is in agricultural production activity at the moment, and canceling the alarm
Advantageous effects
Compared with the prior art, the invention has the following remarkable improvements: 1. the invention automatically detects the condition that the alarm is triggered when the pseudo change occurs in the monitoring area in the traditional fixed-point monitoring and removes the false alarm, thereby solving the problems of low efficiency and high cost of the traditional manual identification and operation. 2. The invention adopts different detection modes aiming at the weather change and the farmland change, thereby improving the detection accuracy. 3. The invention detects the rain/snow scene in the same frame of video image, thus reducing the interference caused by environmental illumination and other environmental factors.
Drawings
FIG. 1 is a flow chart of the steps of an embodiment of the method of the present invention;
FIG. 2 is a flow chart of image rain/snow feature identification;
FIG. 3 is a brightness histogram of a monitored picture;
FIG. 4 is a gray scale image of an image to be detected
FIG. 5 is a graph of the effect of median filtering in an image;
fig. 6 is a diagram showing the effect of the image binarized by the image difference method.
Detailed Description
For the purpose of illustrating the technical solutions disclosed in the present invention in detail, the following description is further provided with reference to the accompanying drawings and the detailed description.
The invention discloses a method for eliminating pseudo-change of fixed-point land video monitoring, which comprises the following steps as shown in figure 1:
step 1: the time of receiving spring ploughing autumn and the ploughing coverage area information of the monitoring area are stored in the database, the ploughing area frame can be selected by a plurality of rectangular frames to replace the occupied area of the actual ploughing, and comparison is carried out after convenience is achieved.
Step 2: and (3) making a brightness histogram of the image in the monitoring area, and taking a brightness value corresponding to the last peak as a contrast parameter, wherein the method specifically comprises the following steps:
2.1: acquiring a screenshot of a monitoring picture under the condition of no rain or snow, and performing hsv channel separation on the acquired image;
2.2: and drawing a brightness histogram of the image to obtain the maximum brightness value P of the image. Because some noise points may exist in the image, the brightness of some pixel points may be close to 0 or 255, and proper filtering processing can be performed on the image before the brightness histogram is performed, so that the obtained P value is more accurate;
2.3: setting a threshold value T to be 0.8P, and counting the number n of pixels with brightness larger than T;
and step 3: when the system gives an alarm, the size of an alarm area is obtained, if the alarm area exceeds 40% of the monitoring area, a next detection mechanism is triggered, and whether the alarm is caused by the false change is judged.
And 4, step 4: judging whether the weather information acquired from the network service is a pseudo change caused by rain/snow according to the weather information acquired from the network service and the characteristics of rain/snow in the image, and specifically comprising the following steps:
4.1: the brightness change of the pixel points affected by raindrops and snowflakes in the image is large. Firstly, graying a current monitoring picture, wherein the change degrees of raindrops and snowflakes to R, G, B components in an experiment are basically the same, so that a gray formula adopts I ═ R + G + B)/3, and a gray graph of an image to be detected is shown in FIG. 4;
4.2: the grayscale image obtained in 4.1 is median filtered, and the filtered image is shown in fig. 5. The snowflake or raindrop-like features present in the image are removed;
4.3: and (3) performing difference between the gray level image of the original image obtained in the step (4.1) and the gray level image after median filtering obtained in the step (4.2), setting a threshold value T, and performing binarization on the image, wherein the image after binarization is shown in a step (6). The value of T in this example is 30. The formula of binarization is as follows:
wherein the content of the first and second substances,
the gray values of the pixel points of the gray image of the original image are obtained;
the gray value of the pixel point of the filtered gray image is obtained; pi,jAnd obtaining the gray level of the pixel point of the binarized image.
4.4: if the total number of pixel points of the image is M x N, if
If the value of (A) is greater than a predetermined threshold value, it is judged that snow/rain is present in the image, and the value calculated in this example is
The value was 5%.
And 5: acquiring weather information in a week, judging whether the snowfall occurs in the week, wherein the week is preferably the first 7 days from the current day of alarm; meanwhile, areas meeting the snow characteristics in the image are extracted, and the ratio of the areas to the monitoring area is used for judging whether snow exists or not, wherein the method specifically comprises the following steps:
5.1: the system acquires local real-time weather information through the Webservice and judges whether snowfall weather occurs within a week.
5.2: and extracting a suspected snow area in the image according to the features of the snow. The component with the largest value among R, G, B three components of a pixel point is recorded as Max, the minimum value is recorded as Min, and when the (Max-Min) < alpha and (R + G + B)/3> beta are met, the pixel point can be judged to meet the color and brightness characteristics of the accumulated snow. The value of alpha is selected according to the contrast, and is generally less than 20. The value of β can be chosen according to the maximum lightness P of the image, typically β ═ 0.8 × P.
Step 5.3: and judging whether the coverage rate of the area meeting the snow cover characteristics to the picture is more than 50%, if so, indicating that the snow cover exists in the monitoring area at the moment.
Step 6: judging whether the local is in a busy farming period or not by comparing the spring ploughing and autumn harvesting time in the database with the real-time of the local, comparing the alarm area with the cultivated land coverage area stored in the step 1, if the alarm area is in the cultivated land area, judging that the alarm area is a false change caused by cultivation, and canceling the alarm;
6.1: and acquiring local real-time through network service, comparing the local real-time with the spring and autumn time in the database, and judging whether the local real-time is in agricultural activities.
6.2, comparing the alarm area with the cultivated land area stored in the step 1, and judging that the monitoring area is in agricultural production activity at the moment if the coincidence rate of the alarm area and the cultivated land area exceeds 95 percent.
Claims (3)
1. A method for eliminating the pseudo-change of fixed-point land video monitoring is characterized by comprising the following steps:
(1) storing the spring ploughing and autumn harvesting time and ploughing coverage area information of the monitoring area into a database;
(2) making a brightness histogram of an original image of the monitoring area, and calculating the range size of the high-brightness area;
(3) when the system gives an alarm, calculating the range of an alarm area;
(4) if the range of the alarm area exceeds 30% of the monitoring area, performing the step (5);
(5) and when at least one of the following three judgments is true, judging the alarm caused by the false change:
(5.1) judging whether snowfall/rain occurs: acquiring weather information from the network service, judging whether the weather information is a pseudo change caused by rain/snow by combining with the characteristics of rain/snow in the image, and if so, canceling the alarm;
(5.2) judging whether snow is accumulated: acquiring weather information in a week, judging whether a snowfall condition occurs in the week, extracting an area meeting the snow characteristics in an image, judging whether a monitored area is covered by snow, and if so, cancelling alarm;
(5.3) judging whether the activity is agricultural production activity: comparing the spring ploughing and autumn harvesting time in the database with the local real-time, judging whether the local is in agricultural production or not by combining the coincidence rate of the alarm area and the ploughing area, and if so, canceling the alarm;
when the three judgments are all 'false', normal alarm is carried out;
the step (2) comprises the following steps:
(2.1) acquiring a screenshot of a monitoring picture under the condition of no rain or snow, and performing hsv channel separation on the acquired image;
(2.2) making a brightness histogram of the image to obtain the maximum brightness value P of the image;
(2.3) setting a threshold value T ═ a × P, wherein a is used for judging high-brightness pixel points in the original image, a is more than or equal to 0.8 and less than or equal to 0.9, counting the number n of pixels with brightness larger than T, and n is the number of the high-brightness pixel points in the original image;
the step (5.1) specifically comprises the following steps:
(5.1.1) graying the current monitoring picture by adopting a gray formula I ═ R + G + B)/3;
(5.1.2) performing median filtering on the gray level image obtained in (5.1.1), and removing the characteristics of snowflakes or raindrops existing in the image;
(5.1.3) making a difference between the gray level images obtained in the step (5.1.1) and the gray level image obtained in the step (5.1.2), setting a threshold value T 'of a brightness difference value, comparing the difference between the brightness values of two pixel points with the threshold value T', judging whether the pixel point meets the raindrop and snowflake characteristics, counting the number n 'of the pixels, and taking 20-40 of the threshold value T';
(5.1.4) if the total number of pixels in the image is M x N, if
If the value of (1) is greater than the threshold value T '', judging that snow/rain exists in the image; the threshold value T '' is a threshold value of the ratio of the pixel point number meeting the snowflake feature to the total pixel point number of the whole image, if the ratio is larger than the threshold value T '', the characteristic pixels with a large number of suspected snowflakes exist in the image, and whether the image snows or not is judged; the threshold value T' takes a value of 3% -5%.
2. A method for eliminating pseudo-variations of fixed-site video surveillance of the earth according to claim 1, characterized in that said step (5.2) comprises in particular the following:
(5.2.1) the system acquires local real-time weather information through the meteorological platform and judges whether snowfall weather occurs in one week;
(5.2.2) extracting a suspected snow area in the image according to the features of the snow; the component with the largest value in R, G, B three components of a pixel point is marked as Max, the minimum value is marked as Min, and when the (Max-Min) < alpha and (R + G + B)/3> beta are met, the pixel point is judged to meet the color and brightness characteristics of the accumulated snow; the value of alpha is selected according to the contrast, and is less than 20; selecting the value of beta according to the maximum lightness P of the image, wherein the value of beta is beta-b-P, b is used for judging high-lightness pixel points in the image, and b is more than or equal to 0.8 and less than or equal to 0.9;
(5.2.3) judging whether the coverage rate of the area meeting the snow characteristics to the picture is greater than s%, and is more than or equal to 20% and less than or equal to 50%, if so, indicating that the snow coverage condition exists in the monitored area at the moment, and stopping alarming; the specific value of s is selected according to the snowfall condition of the local previous year, and the value is larger in the area with more severe snowfall.
3. A method for eliminating pseudo-variations of fixed-site video surveillance of the earth according to claim 1, characterized in that said step (5.3) comprises in particular the following:
(5.3.1) acquiring local real-time through network service, comparing the local real-time with the spring and autumn time in the database, and judging whether the local real-time is in the agricultural production period;
(5.3.2) comparing the alarm area with the cultivated land area stored in the step (1), and if the coincidence rate of the alarm area and the cultivated land area exceeds c percent and c percent is more than or equal to 70 percent and less than or equal to 95 percent, judging that the monitoring area is in agricultural production activity at the moment, and canceling the alarm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810297035.4A CN108734093B (en) | 2018-04-04 | 2018-04-04 | Method for eliminating pseudo change of fixed-point land video monitoring |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810297035.4A CN108734093B (en) | 2018-04-04 | 2018-04-04 | Method for eliminating pseudo change of fixed-point land video monitoring |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108734093A CN108734093A (en) | 2018-11-02 |
| CN108734093B true CN108734093B (en) | 2022-02-11 |
Family
ID=63941224
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810297035.4A Active CN108734093B (en) | 2018-04-04 | 2018-04-04 | Method for eliminating pseudo change of fixed-point land video monitoring |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108734093B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103491351A (en) * | 2013-09-29 | 2014-01-01 | 东南大学 | Intelligent video monitoring method for illegal buildings |
| CN103546726A (en) * | 2013-10-28 | 2014-01-29 | 东南大学 | Method for automatically discovering illegal land use |
| CN105096532A (en) * | 2015-07-30 | 2015-11-25 | 西安石油大学 | Geological disaster predication system |
| CN105512608A (en) * | 2015-11-26 | 2016-04-20 | 北京电信易通信息技术股份有限公司 | Method and system for determining man-made geomorphic feature change |
| US20170308753A1 (en) * | 2016-04-26 | 2017-10-26 | Disney Enterprises, Inc. | Systems and Methods for Identifying Activities and/or Events in Media Contents Based on Object Data and Scene Data |
-
2018
- 2018-04-04 CN CN201810297035.4A patent/CN108734093B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103491351A (en) * | 2013-09-29 | 2014-01-01 | 东南大学 | Intelligent video monitoring method for illegal buildings |
| CN103546726A (en) * | 2013-10-28 | 2014-01-29 | 东南大学 | Method for automatically discovering illegal land use |
| CN105096532A (en) * | 2015-07-30 | 2015-11-25 | 西安石油大学 | Geological disaster predication system |
| CN105512608A (en) * | 2015-11-26 | 2016-04-20 | 北京电信易通信息技术股份有限公司 | Method and system for determining man-made geomorphic feature change |
| US20170308753A1 (en) * | 2016-04-26 | 2017-10-26 | Disney Enterprises, Inc. | Systems and Methods for Identifying Activities and/or Events in Media Contents Based on Object Data and Scene Data |
Non-Patent Citations (2)
| Title |
|---|
| Single image haze removal using integrated dark and bright channel prior;Singh Dilbag 等;《MODERN PHYSICS LETTERS B》;20180210;全文 * |
| 视频大数据分析技术行业应用研究;唐志鸿;《江苏通信》;20180215;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108734093A (en) | 2018-11-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9230175B2 (en) | System and method for motion detection in a surveillance video | |
| CN102348128B (en) | Surveillance camera system having camera malfunction detection function | |
| US6434254B1 (en) | Method and apparatus for image-based object detection and tracking | |
| US20160026865A1 (en) | Vision-based system for dynamic weather detection | |
| CN110148093B (en) | Image defogging improvement method based on dark channel prior | |
| DE19840873A1 (en) | Method and device for automatic forest fire detection | |
| CN101908142A (en) | Feature analysis-based video flame detecting method | |
| CN109359593B (en) | Rain and snow environment picture fuzzy monitoring and early warning method based on image local grid | |
| CN108765453B (en) | Expressway agglomerate fog identification method based on video stream data | |
| WO1998028706B1 (en) | Low false alarm rate video security system using object classification | |
| CN107274410B (en) | Self-adaptive artificial target constant false alarm rate detection method | |
| CN116665080B (en) | Unmanned aerial vehicle deteriorated insulator detection method and system based on target recognition | |
| KR101224494B1 (en) | Video based smoke detection method robust to light conditions | |
| CN114639075A (en) | Method and system for identifying falling object of high altitude parabola and computer readable medium | |
| CN112784738A (en) | Moving object detection alarm method, device and computer readable storage medium | |
| CN111179302A (en) | Moving target detection method and device, storage medium and terminal equipment | |
| CN112800846A (en) | High-altitude parabolic monitoring method and device, electronic equipment and storage medium | |
| CN115311623A (en) | Equipment oil leakage detection method and system based on infrared thermal imaging | |
| CN110691224A (en) | Transformer substation perimeter video intelligent detection system | |
| CN113392754B (en) | Method for reducing false pedestrian detection rate based on yolov5 pedestrian detection algorithm | |
| CN1556506A (en) | Intelligent alarming treatment method of video frequency monitoring system | |
| CN113409360A (en) | High altitude parabolic detection method and device, equipment and computer storage medium | |
| CN112613347B (en) | Automatic recognition method for fire passing range and burning degree of forest fire | |
| CN108734093B (en) | Method for eliminating pseudo change of fixed-point land video monitoring | |
| KR102040562B1 (en) | Method to estimate visibility distance using image information |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |