A building indoor video monitoring system
By combining video acquisition and analysis units in the video surveillance system, the aspiration and conversion analysis of image data are carried out, and the predicted value is calculated to achieve early warning of the number of people in the existing technology, which solves the problem that the number of people entering the building cannot be accurately predicted, and improves the problem of improving the number of people entering the building. The coordination capabilities of the monitoring system are achieved.
Patent Information
- Application Number
- CN202210133767.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-02-14
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2042-02-14
AI Technical Summary
The existing video surveillance system cannot accurately predict the number of people entering the building and cannot effectively coordinate the flow of people.
Design an indoor video surveillance system for building, including video acquisition unit, preliminary analysis unit, binary monitoring unit, processor, repository, early warning rule base, display unit and management unit, and use preliminary analysis by obtaining outdoor and indoor images in real time The unit conducts forward data analysis, combines the binary monitoring unit and processor for conversion analysis, and calculates the predicted value to early warning personnel to coordinate.
It realizes accurate prediction and early warning of the number of people about to enter the building, helps staff to coordinate the flow of people in a timely manner, and improves the accuracy and practicality of the monitoring system.
Smart Images

Figure CN114466168B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of video monitoring, in particular to a building indoor video monitoring system. Background Art
[0002] Monitoring systems have been increasingly widely used. At present, most monitoring equipment on the market can monitor 24 hours a day. The monitor can actively control the monitoring camera to monitor according to the monitor's wishes.
[0003] The Chinese patent with publication number CN101720031A discloses an indoor video monitoring and alarm method based on the differential method. First, an image in a normal state is collected as the background image; then an image is taken at intervals, and the absolute value of the difference between the pixel values of the corresponding points of the image and the background image is calculated to obtain a differential image; each pixel of the differential image is scanned. If the pixel value of a certain pixel is greater than 0, it means that the pixel value here has changed significantly compared with the pixel value of the corresponding point in the background image. If the total number of changed pixels n≥ the preset threshold K, it means that the indoor situation is abnormal at this moment, and the user does not need to pay attention to the screen all the time. For example, Chinese patent CN203942603U discloses an indoor video monitoring device with active warning, and Chinese patent CN107135378A discloses a video monitoring system, both of which provide technical support for indoor security monitoring.
[0004] For example, Chinese patent CN108900812A discloses an indoor video monitoring method based on remote control, Chinese patent CN210168132U discloses a video monitoring system connected to the Internet of Things, and so on. All of these can provide users with early warning information in a timely manner, but they cannot solve the problem of accurately predicting the number of people entering a building. Summary of the invention
[0005] The purpose of the present invention is to provide a building indoor video monitoring system.
[0006] The purpose of the present invention can be achieved through the following technical solutions:
[0007] A building indoor video monitoring system, comprising a video acquisition unit, a preliminary analysis unit, a two-item monitoring unit, a processor, a storage library, an early warning rule library, a display unit, and a management unit;
[0008] The video acquisition unit is a monitoring camera device installed outside the building, which is used to acquire outdoor real-time images in real time and mark them as outdoor images; the video acquisition unit is used to transmit the outdoor images to the preliminary analysis unit;
[0009] The preliminary analysis unit receives the outdoor image transmitted by the video acquisition unit and performs a data analysis on the outdoor image. The specific method of the data analysis is as follows:
[0010] Step 1: Based on the current time, push forward the inertia time. The inertia time is preset by the management personnel, usually half a year or other more easily calculated time;
[0011] Step 2: Taking one day as a unit, obtain the number of people passing through the outdoor of the building on all days, mark it as the number of people coming and going Di,i=1...n; mark the time of the corresponding day as Ui,i=1...n;
[0012] Step 3: Then let i=1, obtain the outdoor image corresponding to that day, perform intention analysis on the outdoor image of that day, and obtain the number of people with intention marks, which is the intention number Y1 and the intention face;
[0013] Step 4: Repeat the process of step 3 by adding one to the value of i, and obtain the number of intentions and the intended faces corresponding to all Di; mark them as Yi, i = 1...n; Yi has a one-to-one correspondence with Di and Ui;
[0014] The preliminary analysis unit is used to transmit all outdoor images within the Di range, the intended number Yi and the incoming and outgoing singular number Di and the corresponding intended human faces to the processor;
[0015] The two monitoring units are arranged indoors in the building, and are used to obtain images of all persons entering the building, and mark them as indoor images. The two monitoring units are used to transmit the indoor images to the processor, and the processor receives the indoor images transmitted by the two monitoring units, and performs conversion analysis to obtain a kernel-to-core conversion value H.
[0016] The preliminary analysis unit is also used to obtain the real-time number of interested persons and transmit the number of interested persons to the processor, which is used to multiply the number of interested persons by the conversion value H to obtain a predicted value.
[0017] Furthermore, the specific method of the intention analysis in step three is:
[0018] S1: Get the corresponding outdoor image;
[0019] S2: Afterwards, the first person who enters the outdoor monitoring area of the building is obtained and marked as a single target;
[0020] S3: Obtain the rear point of the left foot of the single item target. The rear point is determined by the back of the sticker. The back of the sticker is used to gradually stick forward from the back of the single item target. The position where the left foot is initially touched is marked as the rear point.
[0021] S4: Then, the same principle as step S3 is used to obtain the right foot rear point of the single item, and mark it as rear foot point 2;
[0022] S5: Then, the front point of the left foot is obtained in the same way;
[0023] Connect the front point and the back point to get the left line;
[0024] S6: Obtain the front point of the right foot in the manner of step S5, mark it as front point 2, connect the front point 2 and the back point 2 to obtain a right-pointing line;
[0025] S7: Mark the acute angle area formed by the left-pointing line and the right-pointing line as the target direction area;
[0026] S8: Obtain the location of the entrance inside the building, and continuously observe the direction of the single item mark for T1 time, where T1 is a preset value;
[0027] S9: When it is obtained that the intersection of the orientation area and the building room entrance exceeds α degrees within the time T1, it is expressed as intention, the intention time is obtained, α is a preset value, the intention time is divided by T1, and the obtained value is marked as the intention ratio. When the intention ratio exceeds X1, the single item is marked as the intention target; the face of the intention target is obtained and marked as the intention face;
[0028] S10: Perform the same determination of steps S2-S10 on all persons in the outdoor image, and then obtain all intended targets; and synchronously obtain all intended faces;
[0029] S11: Get the number of people who have received the intention mark and mark it as the intention number Y1.
[0030] Furthermore, the back surface in step S3 is the plane with the largest area where the back is located and all planes parallel to it, which are determined in this way.
[0031] Furthermore, the method for obtaining the forefoot point in step S5 is to obtain the sole of the left foot and the forefoot point of the third toe at the same time, where the forefoot point is the frontmost point of the toe, and connect the sole and the forefoot point to obtain a target line;
[0032] Merge the marking line parallel to the back point of the left foot, and then mark the front point of the left foot as the front point.
[0033] Furthermore, the specific method of conversion analysis is as follows:
[0034] SS1: Get the indoor image corresponding to the day of Di according to the timestamp;
[0035] SS2: Compare the indoor image with the outdoor image, obtain the faces of people who enter the indoor image and the outdoor image, and mark the faces that are consistent as the number of people entering; obtain the faces of the people entering and mark them as the faces of the people entering;
[0036] SS3: Let i=1, obtain the incoming face corresponding to the day where U1 is located, compare the incoming face with the intended face corresponding to the day where D1 is located, mark the consistent ones as the converted face, and divide the converted face by the intended face to get the mark as the conversion score ratio;
[0037] SS4: Repeat the process of adding 1 to the value of i and repeat step SS3 to obtain the conversion ratio of all Ui, which is marked as Zi; Zi and Ui are in one-to-one correspondence;
[0038] SS5: Then the mean of Zi is automatically calculated and marked as P; then the deviation value L is automatically calculated using the formula. The specific calculation formula is:
[0039]
[0040] SS6: When L exceeds X2, a reduction signal is generated. When the reduction signal is generated, the values of Zi are deleted in descending order according to |Zi-P|. Each time a Zi is deleted, L is calculated once by repeating step SS5.
[0041] SS7: Repeat the determination of step SS6 until no nuclear subtraction signal is generated, calculate the mean of the remaining Zi, and mark it as the nuclear conversion value H.
[0042] Furthermore, the processor is used to perform early warning analysis on the predicted value in combination with the early warning rule library. The specific method of early warning analysis is:
[0043] When the predicted value exceeds X3, an early warning signal is generated; X3 is a preset value;
[0044] The processor is used to display "too many people at present, please pay attention to coordination" in real time through the display unit when generating the warning signal;
[0045] The processor is used to transmit the core-to-core conversion value H and the predicted value to the display unit for real-time display; the processor is used to transmit the core-to-core conversion value H and the predicted value to the storage repository for real-time storage and overwrite the original value.
[0046] Furthermore, the management unit is communicatively connected with the processor for inputting all preset values.
[0047] Furthermore, the video acquisition unit, the preliminary analysis unit and the binomial monitoring unit perform the prediction process of the conversion value H and the prediction value once at fixed intervals.
[0048] Beneficial effects of the present invention:
[0049] The present invention obtains outdoor real-time images in real time through a video acquisition unit and marks them as outdoor images; then uses a preliminary analysis unit to obtain all outdoor images within the Di range, the intention number Yi and the single number Di and the corresponding intention face through analysis; then uses a two-item monitoring unit and a processor to perform conversion analysis to obtain a core-to-core conversion value H; analyzes the conversion value through past data, uses the conversion value to reasonably obtain a predicted value, and selects whether to remind the staff to pay attention to coordination according to the predicted value, so as to accurately predict the number of people who are about to enter the building. The present invention is simple, effective, and easy to use. BRIEF DESCRIPTION OF THE DRAWINGS
[0050] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.
[0051] Figure 1 The system block diagram of the building indoor video monitoring system of the present invention. DETAILED DESCRIPTION
[0052] like Figure 1 As shown, a building indoor video monitoring system includes a video acquisition unit, a preliminary analysis unit, a two-item monitoring unit, a processor, a storage library, an early warning rule library, a display unit, and a management unit;
[0053] The video acquisition unit is a monitoring camera device installed outside the building, which is used to acquire outdoor real-time images in real time and mark them as outdoor images; the video acquisition unit is used to transmit the outdoor images to the preliminary analysis unit;
[0054] The preliminary analysis unit receives the outdoor image transmitted by the video acquisition unit and performs a data analysis on the outdoor image. The specific method of the data analysis is as follows:
[0055] Step 1: Based on the current time, push forward the inertia time. The inertia time is preset by the management personnel, usually half a year or other more easily calculated time;
[0056] Step 2: Taking one day as a unit, obtain the number of people passing through the outdoor of the building on all days, and mark it as the number of people passing by Di,i=1...n; mark the time of the corresponding day as Ui,i=1...n; this technology is an existing technology, and it is easy for those skilled in the art to implement the technology, so it will not be described in detail;
[0057] Step 3: Then let i=1, obtain the outdoor image corresponding to that day, and perform intention analysis on the outdoor image of that day. The specific method of intention analysis is:
[0058] S1: Get the corresponding outdoor image;
[0059] S2: Afterwards, the first person who enters the outdoor monitoring area of the building is obtained and marked as a single target;
[0060] S3: Obtain the rear point of the left foot of the single item target. The rear point is determined by the back of the sticker. The back of the sticker is used to gradually stick forward from the back of the single item target. The position where the left foot is initially touched is marked as the rear point.
[0061] The back side of the back is the plane with the largest area of the back and all planes parallel to it, which are determined in this way;
[0062] S4: Then, the same principle as step S3 is used to obtain the right foot rear point of the single item, and mark it as rear foot point 2;
[0063] S5: Then, the front point of the left foot is obtained in the same way;
[0064] The method of obtaining the front point of the foot is to obtain the sole of the left foot and the front point of the third toe at the same time. The front point of the toe is the front point of the toe. The sole of the foot and the front point of the toe are connected to obtain the target line.
[0065] Merge the marked line parallel to the back point of the left foot, and then mark the front point of the left foot as the front point;
[0066] Connect the front point and the back point to get the left line;
[0067] S6: Obtain the front point of the right foot in the manner of step S5, mark it as front point 2, connect the front point 2 and the back point 2 to obtain a right-pointing line;
[0068] S7: Mark the acute angle area formed by the left-pointing line and the right-pointing line as the target direction area;
[0069] S8: Obtain the location of the entrance inside the building, and continuously observe the direction of the single item mark for T1 time, where T1 is a preset value;
[0070] S9: When it is obtained that the intersection of the orientation area and the building room entrance exceeds α degrees within the time T1, it is expressed as intention, the intention time is obtained, α is a preset value, the intention time is divided by T1, and the obtained value is marked as the intention ratio. When the intention ratio exceeds X1, the single item is marked as the intention target; the face of the intention target is obtained and marked as the intention face;
[0071] S10: Perform the same determination of steps S2-S10 on all persons in the outdoor image, and then obtain all intended targets; and synchronously obtain all intended faces;
[0072] S11: Get the number of people who have obtained the intention mark, and mark it as the intention number Y1;
[0073] Step 4: Repeat the process of step 3 by adding one to the value of i, and obtain the number of intentions and the intended faces corresponding to all Di; mark them as Yi, i = 1...n; Yi has a one-to-one correspondence with Di and Ui;
[0074] The preliminary analysis unit is used to transmit all outdoor images within the Di range, the intended number Yi and the incoming and outgoing singular number Di and the corresponding intended human faces to the processor;
[0075] The two monitoring units are arranged indoors in the building, and are used to obtain images of all persons entering the building, and mark them as indoor images. The two monitoring units are used to transmit the indoor images to the processor, and the processor receives the indoor images transmitted by the two monitoring units and performs conversion analysis. The specific method of conversion analysis is as follows:
[0076] SS1: Get the indoor image corresponding to the day of Di according to the timestamp;
[0077] SS2: Compare the indoor image with the outdoor image, obtain the faces of people who enter the indoor image and the outdoor image, and mark the faces that are consistent as the number of people entering; obtain the faces of the people entering and mark them as the faces of the people entering;
[0078] SS3: Let i=1, obtain the incoming face corresponding to the day where U1 is located, compare the incoming face with the intended face corresponding to the day where D1 is located, mark the consistent ones as the converted face, and divide the converted face by the intended face to get the mark as the conversion score ratio;
[0079] SS4: Repeat the process of adding 1 to the value of i and repeat step SS3 to obtain the conversion ratio of all Ui, which is marked as Zi; Zi and Ui are in one-to-one correspondence;
[0080] SS5: Then the mean of Zi is automatically calculated and marked as P; then the deviation value L is automatically calculated using the formula. The specific calculation formula is:
[0081]
[0082] SS6: When L exceeds X2, a reduction signal is generated. When the reduction signal is generated, the values of Zi are deleted in descending order according to |Zi-P|. Each time a Zi is deleted, L is calculated once by repeating step SS5.
[0083] SS7: Repeat the determination of step SS6 until no nuclear reduction signal is generated, calculate the mean of the remaining Zi, and mark it as the nuclear conversion value H;
[0084] The preliminary analysis unit is also used to obtain the real-time number of intended persons and transmit the number of intended persons to the processor, and the processor is used to multiply the number of intended persons by the conversion value H to obtain a predicted value;
[0085] The video acquisition unit, the preliminary analysis unit and the binomial monitoring unit perform the prediction process of the conversion value H and the prediction value once every fixed period;
[0086] The processor is used to perform early warning analysis on the predicted value in combination with the early warning rule library. The specific method of early warning analysis is as follows:
[0087] When the predicted value exceeds X3, an early warning signal is generated; X3 is a preset value;
[0088] The processor is used to display "too many people at present, please pay attention to coordination" in real time through the display unit when generating the warning signal;
[0089] The management unit is in communication with the processor and is used to input all preset values;
[0090] The processor is used to transmit the nuclear conversion value H and the predicted value to the display unit for real-time display; the processor is used to transmit the nuclear conversion value H and the predicted value to the storage library for real-time storage and overwrite the original value;
[0091] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.
Claims
1. A building indoor video monitoring system, characterized in that: include: Video acquisition unit: It is a surveillance camera device installed outside the building, used to acquire real-time outdoor images in real time, and transmit them to the preliminary analysis unit after marking them as outdoor images; Preliminary analysis unit: It receives the outdoor image transmitted by the video acquisition unit and performs directional data analysis on the outdoor image. The specific method of directional data analysis is as follows: Step 1: Based on the current time, calculate the inertia time forward. The inertia time is preset by the management personnel. Step 2: Taking one day as a unit, obtain the number of people passing through the outdoor of the building on all days, mark it as the number of people coming and going Di,i=1...n; mark the time of the corresponding day as Ui,i=1...n; Step 3: Then let i=1, obtain the outdoor image corresponding to that day, perform intention analysis on the outdoor image of that day, and obtain the intention number Y1 and the intention face. The number of the intention target is the intention number; Step 4: Repeat the process of step 3 by adding one to the value of i, and obtain the number of intentions and the intended faces corresponding to all Di; mark them as Yi, i = 1...n; Yi has a one-to-one correspondence with Di and Ui; Preliminary analysis unit: It is used to transmit all outdoor images, intention numbers Yi, incoming and outgoing singular numbers Ui and corresponding intention faces within the range of Ui to the processor; Binary monitoring unit: It is set up in the building, used to obtain images of all people entering the building, and mark them as indoor images and transmit them to the processor. The processor receives the indoor images transmitted by the binary monitoring unit, and performs conversion analysis to obtain the kernel conversion value H; The preliminary analysis unit is also used to obtain the real-time number of intended persons and transmit the number of intended persons to the processor, and the processor is used to multiply the number of intended persons by the conversion value H to obtain a predicted value; The specific method of intention analysis in step 3 is as follows: S1: Get the corresponding outdoor image; S2: Afterwards, the first person who enters the outdoor monitoring area of the building is obtained and marked as a single target; S3: Obtain the rear point of the left foot of the single item target. The rear point is determined by the back of the sticker. The back of the sticker is used to gradually stick forward from the back of the single item target. The position where the left foot is initially touched is marked as the rear point. S4: Then, the same principle as step S3 is used to obtain the right foot rear point of the single item, and mark it as rear foot point 2; S5: Then, the front point of the left foot is obtained in the same way; Connect the front point and the back point to get the left line; S6: Obtain the front point of the right foot in the manner of step S5, mark it as front point 2, connect the front point 2 and the back point 2 to obtain a right-pointing line; S7: Mark the acute angle area formed by the left-pointing line and the right-pointing line as the target direction area; S8: Obtain the location of the entrance inside the building, and continuously observe the direction of the single item mark for T1 time, where T1 is a preset value; S9: When the intersection of the orientation area and the building room entrance exceeds α degrees within the time T1, it is expressed as intention, and the intention time is obtained. α is a preset value. The intention time is divided by T1, and the obtained value is marked as the intention ratio. When the intention ratio exceeds X1, the single item is marked as the intention target; the face of the intention target is obtained and marked as the intention face; S10: Perform the same determination of steps S2-S9 on all persons in the outdoor image, and then obtain all intended targets; and synchronously obtain all intended faces; S11: Get the number of intended targets and mark it as intended number Y1; The back surface in step S3 is the plane with the largest area of the back and all planes parallel to it, which are determined in this way; The specific methods of conversion analysis are as follows: SS1: Get the indoor image corresponding to the day of Di according to the timestamp; SS2: Compare the indoor image with the outdoor image, obtain the faces of people who enter the indoor image and the outdoor image, and mark the faces that are consistent as the number of people entering; obtain the faces of the people entering and mark them as the faces of the people entering; SS3: Let i=1, obtain the incoming faces corresponding to the day where U1 is located, compare the incoming faces with the intended faces corresponding to the day where D1 is located, mark the consistent faces as converted faces, and divide the number of converted faces by the number of intended faces to get the conversion ratio; SS4: Repeat the process of adding 1 to the value of i and repeat step SS3 to obtain the conversion ratio of all Ui, which is marked as Zi; Zi and Ui are in one-to-one correspondence; SS5: Then the mean of Zi is automatically calculated and marked as P; then the deviation value L is automatically calculated using the formula. The specific calculation formula is: SS6: When L exceeds X2, a reduction signal is generated. When the reduction signal is generated, the values of Zi are deleted in descending order according to |Zi-P|. Each time a Zi is deleted, L is calculated once by repeating step SS5. SS7: Repeat the determination of step SS6 until no nuclear subtraction signal is generated, calculate the mean of the remaining Zi, and mark it as the nuclear conversion value H.
2. A building indoor video monitoring system according to claim 1, characterized in that: The processor is used to perform early warning analysis on the predicted value in combination with the early warning rule library. The specific method of early warning analysis is as follows: When the predicted value exceeds X3, an early warning signal is generated; X3 is a preset value; The processor is used to display "too many people at present, please pay attention to coordination" in real time through the display unit when generating the warning signal; The processor is used to transmit the core-to-core conversion value H and the predicted value to the display unit for real-time display; the processor is used to transmit the core-to-core conversion value H and the predicted value to the storage repository for real-time storage and overwrite the original value.
3. A building indoor video monitoring system according to claim 1, characterized in that: It also includes a management unit; the management unit is connected to the processor for inputting all preset values.
4. A building indoor video monitoring system according to any one of claims 1 to 3, characterized in that: The video acquisition unit, the preliminary analysis unit and the binomial monitoring unit perform the prediction process of the conversion value H and the prediction value once every fixed period.
