Video geographic positioning method and device based on iron tower monitoring

Abstract

The invention discloses a video geographic positioning method and device based on iron tower monitoring, and the method comprises the steps: constructing a monitoring video homonymy point sample library, collecting the homonymy point data of a monitoring video and a remote sensing orthoimage, and storing a reference image; establishing an initial calibration mapping relation, and realizing coordinate transformation of the remote sensing image and the video image through projection transformation; matching a reference image closest to the real-time video frame from a reference image library through image retrieval; dense matching of inclined images is carried out by using a deep learning model, and automatic alignment of feature points is realized; and resolving a video image coordinate to a geographic coordinate through homography transformation and an initial calibration model to complete high-precision positioning. According to the method, the bidirectional conversion function of finding the land by the video and finding the video by the land is supported, the application efficiency of iron tower monitoring in the fields of cultivated land protection and the like is effectively improved, a good positioning effect is still achieved for complex scenes such as plains, mountainous regions and the like, and corresponding technical support is provided for wide application of iron tower video monitoring.

Description

Technical Field

[0001] The present invention belongs to the field of computer vision, and in particular relates to a video geo-positioning solution based on tower monitoring. Background Art

[0002] In recent years, my country has placed increasing demands on farmland protection. However, traditional satellite imagery, with its inherent lag, has become increasingly inadequate for timely monitoring of farmland. This is evident in the fact that by the time farmland violations are discovered through image changes, they are often already established, with substantial damage to the cultivated layer beyond repair, increasing the difficulty and cost of addressing them.

[0003] With technological advancements and the rise of high-altitude video, tower video has been introduced into farmland protection efforts. Automatic positioning technology is a crucial aspect of video technology, and precise positioning is essential for widespread and in-depth application. Mastering automatic positioning technology will not only effectively improve monitoring efficiency and reduce labor input, but also mitigate climate change impacts. Therefore, its application across various industries is expected to expand. Currently, research on automatic positioning technology and positioning accuracy for tower video is still in its infancy. While some research has been conducted in various fields, most of this research focuses on low-altitude video, fixed focal length, and fixed orientation. Limited research has focused on the automatic positioning of targets monitored by high-altitude tower video. Due to the constant rotation of high-altitude cameras and the influence of focal length variations, video images are distorted, making it difficult to align them with plane coordinates, making accurate positioning impossible using high-definition cameras. If the automatic positioning technology of tower video can be solved and the relevant locations can be accurately pushed to the staff, it will be beneficial for grassroots staff to "target" their inspections on farmland protection, greatly improve the efficiency and accuracy of inspections, and help "early detection and early prevention" of clues of farmland violations, thereby better promoting the in-depth and practical work of farmland protection.

[0004] The integration of intelligent video surveillance and geographic information system data offers new avenues for natural resource monitoring and protection. Efficient cross-modal data integration requires aligning video images with map coordinates. Currently, some provincial-level tower projects in my country have initially achieved 24 / 7 dynamic monitoring of key natural resource conservation areas across the province. However, accurate and rapid acquisition of the geographic coordinates of interesting features in video images remains a challenge. Aligning video images with orthophoto map coordinates facilitates rapid identification of the geographic locations of areas of interest, enabling research such as buffer zone analysis and influencing factor assessment. Therefore, solutions are urgently needed in this area. Summary of the Invention

[0005] In view of the shortcomings of existing technologies and combining the data characteristics of tower monitoring videos, this patent provides a high-precision video geolocation technology solution based on tower monitoring.

[0006] To achieve the above object, the technical solution of the present invention is a video geolocation method based on tower monitoring, comprising the following steps:

[0007] Construct a sample library of surveillance video homonymous points, including collecting homonymous point data of surveillance videos and remote sensing orthophotos, and saving surveillance images as reference images;

[0008] Constructing a preliminary calibration mapping relationship, including establishing a preliminary calibration model between remote sensing images and video images using samples of the same name points;

[0009] Implementing the image retrieval process, including retrieving the reference image closest to the real-time video frame from the reference image library;

[0010] Perform dense matching of oblique images, including automatic matching of feature points between real-time video frames and reference images through deep learning models;

[0011] Perform video geographic coordinate solution, including converting the matched image coordinates into geographic coordinates through the initial calibration model;

[0012] Realize location finding by video, including converting the area of interest in real-time video into geographic coordinates and locating it in remote sensing images.

[0013] Furthermore, the steps of constructing a surveillance video homonymous point sample library include:

[0014] Obtain the surveillance video stream address and camera center point position through the device code and channel code;

[0015] Control the monitoring PTZ to rotate to the specified position and save the image. The image name includes the rotation value, pitch value and focal length;

[0016] Select the points with the same name as the monitoring image from the remote sensing orthophoto and record the three-dimensional image coordinates and geographic coordinates.

[0017] Furthermore, the step of constructing the initial calibration mapping relationship includes:

[0018] The geographic coordinates are converted into calibrated spatial coordinates using a projection transformation formula, in which the rotation angle and center point position parameters of the camera are introduced.

[0019] Furthermore, the steps of the image retrieval process include:

[0020] The closest reference image is retrieved from the reference image library according to the rotation value of the real-time video frame.

[0021] Furthermore, the step of dense matching of the inclined images comprises:

[0022] Use convolutional neural networks to extract local features and generate multi-level feature maps;

[0023] Reconstruct features through positional encoding and Transformer model;

[0024] Perform coarse-grained matching, calculate the correlation between feature points and select high-confidence matching points;

[0025] Perform fine-grained matching, refine the coarse matching to a fine-level feature map, and obtain matching results with sub-pixel accuracy.

[0026] Moreover, the step of calculating the video geographic coordinates includes:

[0027] Mapping matching points from real-time video frames to reference image coordinates through homography transformation;

[0028] The image coordinates are converted into geographic coordinates by combining the initial calibration model.

[0029] In addition, it also includes the steps to implement the function of finding videos by location:

[0030] Select the area of interest in the remote sensing image and calculate the geographic coordinates of its center point;

[0031] Find the reference image with the closest geographic coordinates through image retrieval;

[0032] The geographic coordinates are inversely solved into image coordinates and projected into the real-time video frame.

[0033] On the other hand, the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the video geo-positioning method based on tower monitoring as described above is implemented.

[0034] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the video geo-positioning method based on tower monitoring as described above.

[0035] On the other hand, the present invention also provides a computer program product, including a computer program, characterized in that when the computer program is executed by a processor, the video geo-positioning method based on tower monitoring as described above is implemented.

[0036] The beneficial effects of the technical solution provided by the present invention are:

[0037] (1) The present invention establishes a corresponding coordinate relationship between real-time monitoring video images and remote sensing orthophoto images, achieving high-precision positioning between video and geography. Furthermore, the present invention supports the bidirectional automatic conversion of tower video and geography coordinates from the tower monitoring video system, enabling “finding video by geography” and “finding geography by video”.

[0038] (2) The present invention achieves positioning accuracy of approximately 10 meters in plain areas and approximately 30 meters in hilly areas within the current monitoring range of a tower within one kilometer. It also maintains good positioning results in complex scenarios such as plains and mountains, and the method is highly robust. The high-precision geolocation method proposed in this invention provides corresponding technical support for the widespread application of tower video monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 Flowchart of an embodiment of the present invention.

[0040] Figure 2 This is a structural diagram of the tilted image dense matching network model according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The technical solution of the present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0042] Example 1

[0043] The embodiment of the present invention proposes a video geo-positioning method based on tower monitoring. First, the monitoring image and the remote sensing orthophoto are marked with the same-name points according to the parameters such as the rotation angle, pitch angle and focal length of the PTZ camera to obtain sufficient samples of the same-name points. These same-name points are used to establish an initial calibration model, and the monitoring video image is automatically stored in the reference sample library. Subsequently, an oblique image dense matching model is constructed to realize the automatic matching of the feature points of the ground objects in different monitoring video frames. Finally, the image coordinates of the real-time video frame are passed through the corresponding matching reference samples to realize the process of converting the coordinates into geographic coordinates. The process mainly includes four processes: marking the same-name points, initial calibration, dense matching of oblique images, and video geographic coordinate solution.

[0044] See also Figure 1 The embodiment proposes a video geolocation method based on tower monitoring, and the specific implementation process is as follows: Step a, constructing a monitoring video homonymous point sample library.

[0045] The data required for the present invention to collect mainly consists of homonymous point data from surveillance videos and remote sensing orthophotos. For the same surveillance, the monitoring area of interest is determined by selecting P, T, and Z values, and the approximate remote sensing image range is located using the geographic coordinates of the current surveillance camera. The same feature points of the corresponding features in the video surveillance range and the remote sensing orthophoto are considered as a pair of homonymous point samples. While selecting a sufficient number of homonymous point samples, the surveillance image needs to be simultaneously saved and stored as a reference image.

[0046] During specific implementation, it is preferably recommended that in step a, the unique device code and unique channel code of the current device passed in by the service be used to request the video stream RTSP address of the background service. After selecting a suitable dot modeling scenario, it is necessary to pass the PTZ command to the front-end service to control the monitoring PTZ to the specified position, mark and save the same feature points of the ground objects in the image coordinates of the monitoring video and the geographic coordinates of the remote sensing image, and save the image of the monitoring video as a reference image. The saved points with the same name should be in json format, and the image should be in jpg format. The naming rule is "P{rotation value}T{pitch value}Z{focal length}".

[0047] The specific process is as follows:

[0048] 1. Obtain a token by requesting the service, and then obtain the device code and channel code;

[0049] 2. Request the current monitored rtsp video stream address through the device code and channel code, and obtain the center point position c of the monitoring camera x ,c y ,c z ;

[0050] 3. Use the opencv library to read the video stream address, and control the rotation of the monitor according to the specified P, T, and Z values, and save the screenshot. The image is in jpg format and the naming rule is "P{rotation value}T{pitch value}Z{focal length}".

[0051] 4. Obtain a map covering a two-kilometer radius of the health camera's center point from the service by monitoring the camera's center point location.

[0052] 5. Select the appropriate ground feature point in the current monitoring screenshot and add the 3D image coordinates Ref x ,Ref y ,Ref z Record in the json file, and select the coordinates of the corresponding point on the map. The coordinates are geographic coordinates geo x ,geo y ,geo z, the corresponding records are in the json file, with a total of at least 4 pairs of points with the same name, and the file is named "dev_codes.json".

[0053] For the convenience of implementation reference, the database save path is shown as follows using the device code 43070300831327000004:

[0054] ——43070300831327000004 (folder name)

[0055] ————P{rotation value}T{pitch value}Z{focal length}.jpg

[0056] dev_codes.json

[0057] Step b: constructing an initial calibration mapping relationship.

[0058] For the current surveillance camera, the remote sensing image and the video image are initially calibrated by establishing a corresponding mapping relationship between the same-name point samples saved in step a through the reference image.

[0059] In step b, the idea of projection transformation is introduced into the establishment of the initial calibration model. The specific formula is:

[0060]

[0061] In the formula: Ref x ,Ref y ,Ref z is the calibrated spatial coordinate, geo x ,geo y ,geo z is the three-dimensional geographic space coordinate point before calibration, c x ,c y ,c z is the center point of the surveillance camera, θ x ,θ y ,θ z The specific process is to retrieve the device code of the current monitoring from the database, call in the data of the same-name point of the monitoring for specific calculation, and thus establish the initial calibration model.

[0062] Step c: implement the image retrieval process.

[0063] When the real-time video frame of the surveillance video is used as input, it is necessary to perform image retrieval on the reference images in the image library in the hope of finding the reference image closest to the current video frame in preparation for subsequent dense matching.

[0064] In step c, the image retrieval process involves searching the reference image library to find the reference image that most closely matches the live video frame. The retrieval rule uses the P value of the current video frame to search against the P values of all image file names in the reference image library, finding the reference image with the closest P value to the current video frame.

[0065] Step d: Dense matching of tilted images.

[0066] After retrieving the image to obtain the reference image closest to the current video frame, a deep learning model is constructed to achieve the dense matching process. The network model structure diagram is shown in the figure below. Figure 2 As shown, the specific process is as follows:

[0067] 1. Extract local features using CNN

[0068] Convolutional neural networks (CNNs) possess unique inductive bias properties, such as locality and translation equivariance, making them particularly adept at capturing local feature information in images. In the model of this embodiment, a standard convolutional architecture with integrated FPN (a branch of a convolutional neural network) is first employed to extract multi-level, rich feature descriptions from the two input images to be processed. Through a series of operations, a coarse-level feature map with 1 / 8 the dimension of the original image and a fine-level feature map with 1 / 2 the dimension of the original image are obtained.

[0069] 2. Reconstructing feature modules

[0070] This process mainly processes the coarse-level features of 1 / 8 of the original image dimension extracted in the first step, mainly including position encoding and feature reconstruction using the Transformer model.

[0071] First, the local feature map obtained in the first step is expanded into a long sequence and positionally encoded using the following formula. The positional encoding provides unique position information for each element in sine and cosine format. Finally, the positional encoding is added to the feature map for subsequent input into the Transformer model for further processing.

[0072]

[0073] Where pos refers to the position of the current region in the sequence, m refers to the index value of each value in the vector, the formula uses the sine function to encode the even position and the cosine function to encode the odd position, PE(,) is the encoding of the current region, d model is the embedding dimension of the model.

[0074] The Transformer model used in this module includes a self-attention layer and a cross-attention layer. First, the self-attention layer is entered to calculate the relationship between a certain region and all regions in the sequence. Then, the cross-attention layer is entered to calculate the relationship between a certain region and all regions in another sequence. All regions are processed in a loop, and finally the obtained features are visualized in RGB space after PCA dimensionality reduction.

[0075] 3. Coarse-grained matching

[0076] The coarse-grained matching process is oriented towards the entire image, hoping to find mutually matching areas through global search. Specifically, it includes two processes: calculating correlation and normalization, and threshold screening. First, the score matrix between features is defined as S(i, j). The score matrix is calculated by the inner product of the feature vectors, which is specifically expressed as the formula:

[0077]

[0078] in, represents the eigenvector of the i-th point in the left image, represents the feature vector of the j-th point in the right image, τ represents a scaling factor used to scale the similarity score, and i, j in (i, j) represent the i-th point in the left image and the j-th point in the right image respectively.

[0079] Then the score matrix S(i,j) is normalized using the softmax function, and the inner product of the score matrix is used to obtain the matching probability P c (i,j), specifically expressed as the formula:

[0080] P c (i, j) = softmax(S(i, )) j softmax(S(·,j)) i

[0081] Among them, S(i,·) represents the similarity score between the i-th point in the left image and all points in the right image, softmax(S(i,·)) j represents the normalized result of the similarity score between the j-th point in the right image and the i-th point in the left image from the perspective of the left image, S(·,j) represents the similarity score between the j-th point in the right image and all points in the left image, and softmax(S(·,j)) i It represents the normalized similarity score between the i-th point in the left image and the j-th point in the right image from the perspective of the right image.

[0082] Based on the confidence matrix P c , the embodiment selects the confidence level higher than the threshold θ cThe match is taken as the correct match, and the mutual nearest neighbor standard is further used to filter out possible abnormal rough matches. The process is expressed as formula:

[0083]

[0084] in, middle Respectively represent the left The point on the right is Points, Indicates the left picture The point on the right is The matching probability between points, MNN(P c ) indicates that the nearest neighbor distance method is used to select the correct matching point pairs, M c Indicates the correct matching point pairs found by coarse-grained matching.

[0085] 4. Fine-grained matching

[0086] After establishing the coarse matches, the coarse-to-fine matching strategy is used to refine these matches to the fine-level feature map with 1 / 2 the dimension of the original image. For each coarse match The embodiment first locates its position on the fine-level feature map with 1 / 2 the dimension of the original image Then, two groups of local windows of size w×w (w is generally recommended to be 5) are cut, and then N features in each window are processed in a new attention module. f Transform twice to obtain two new local feature vectors after transformation and Respectively and Then, calculate The center point vector and The correlation of each vector in the graph is calculated by the center point of graph A and the w in graph B. 2 The correlation of the points is normalized to generate a probability distribution graph, which represents Each pixel and point in the neighborhood By calculating the expectation of the probability distribution, the final position with sub-pixel accuracy on the B map is obtained. Expressed as Step e: Calculate the geographic coordinates of the video.

[0087] After completing the initial calibration process of step b and the tilted image matching process of step d, the matching points obtained after the tilted image matching are subjected to homography transformation to achieve the image coordinate solution of the region of interest in the current video frame to the reference image. Then, the geographic coordinates corresponding to the region of interest are solved using the mapping relationship constructed previously. The mutual solution between the coordinates needs to be realized in combination with step b and the homography transformation matrix. The specific calculation formula is as follows

[0088]

[0089] In the formula, a1, a2, ..., a9 are conversion parameters, geo x ,geo y ,geo z is the geographic space coordinate, c x ,c y ,c z is the location coordinate of the surveillance camera, θ x ,θ y ,θ z is the rotation angle of the surveillance camera in the x, y, z directions, image x ,image y ,image z The three-dimensional pixel coordinates of the current video frame image.

[0090] Step f is implemented using the video location search function.

[0091] After completing the video geographic coordinate solution in step e, the selected area of interest (image coordinates) of the current real-time video is found in the remote sensing image according to the geographic coordinates obtained by the solution in the above steps and drawn to realize the function of finding the ground by video.

[0092] Example 2

[0093] Based on the steps ae of Example 1, a function of finding a video by location is further proposed.

[0094] Similar to step f, the process of finding the video by geolocation involves selecting an area of interest in the remote sensing image, calculating the geographic coordinates of the area's center point, and using image retrieval to find the reference image with the closest geographic coordinates. Step e then performs the inverse calculation from geographic coordinates to image coordinates, projecting the image coordinates onto the real-time video frame to achieve the video by geolocation functionality.

[0095] Through the above steps, the present invention establishes a corresponding coordinate relationship between the real-time monitoring video image and the remote sensing orthophoto image, realizes high-precision positioning between the video and the geography, and solves the two core functions of finding the ground with the video and finding the video with the ground.

[0096] In specific implementation, the method proposed in the technical solution of the present invention can be automatically run by those skilled in the art using computer software technology. System devices that implement the method, such as computer-readable storage media that store the corresponding computer program of the technical solution of the present invention and computer equipment that runs the corresponding computer program, should also be within the scope of protection of the present invention.

[0097] The video geolocation electronic device based on tower monitoring provided by the present invention is described below. The video geolocation electronic device based on tower monitoring described below and the video geolocation method based on tower monitoring described above can refer to each other.

[0098] The electronic device may include a processor, a communications interface, a memory, and a communications bus, wherein the processor, the communications interface, and the memory communicate with each other via the communications bus. The processor may invoke logic instructions in the memory to execute the video geolocation method based on tower monitoring, which primarily includes the software processing portion described above.

[0099] In addition, the logical instructions in the above-mentioned memory can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0100] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the software processing part of the video geolocation method based on tower monitoring provided by the above-mentioned methods.

[0101] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the software processing part of the video geolocation method based on tower monitoring provided by the above methods.

[0102] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0103] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0104] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A video geolocation method based on tower monitoring, characterized in that: The following steps are involved: Construct a sample library of surveillance video homonymous points, including collecting homonymous point data of surveillance videos and remote sensing orthophotos, and saving surveillance images as reference images; Constructing a preliminary calibration mapping relationship, including establishing a preliminary calibration model between remote sensing images and video images using samples of the same name points; Implementing the image retrieval process, including retrieving the reference image closest to the real-time video frame from the reference image library; Perform dense matching of oblique images, including automatic matching of feature points between real-time video frames and reference images through deep learning models; Perform video geographic coordinate solution, including converting the matched image coordinates into geographic coordinates through the initial calibration model; Realize location finding by video, including converting the area of interest in real-time video into geographic coordinates and locating it in remote sensing images.

2. The video geolocation method based on tower monitoring according to claim 1, characterized in that: The steps of constructing a surveillance video homonymous point sample library include: Obtain the surveillance video stream address and camera center point position through the device code and channel code; Control the monitoring PTZ to rotate to the specified position and save the image. The image name includes the rotation value, pitch value and focal length; Select the points with the same name as the monitoring image from the remote sensing orthophoto and record the three-dimensional image coordinates and geographic coordinates.

3. The video geolocation method based on tower monitoring according to claim 1, characterized in that: The step of constructing the initial calibration mapping relationship includes: The geographic coordinates are converted into calibrated spatial coordinates using a projection transformation formula, in which the rotation angle and center point position parameters of the camera are introduced.

4. The video geolocation method based on tower monitoring according to claim 1, characterized in that: The steps of the image retrieval process include: The closest reference image is retrieved from the reference image library according to the rotation value of the real-time video frame.

5. The video geolocation method based on tower monitoring according to claim 1, characterized in that: The step of dense matching of the inclined images comprises: Use convolutional neural networks to extract local features and generate multi-level feature maps; Reconstruct features through positional encoding and Transformer model; Perform coarse-grained matching, calculate the correlation between feature points and select high-confidence matching points; Perform fine-grained matching, refine the coarse matching to a fine-level feature map, and obtain matching results with sub-pixel accuracy.

6. The video geolocation method based on tower monitoring according to claim 1, characterized in that: The steps of calculating the video geographic coordinates include: Mapping matching points from real-time video frames to reference image coordinates through homography transformation; The image coordinates are converted into geographic coordinates by combining the initial calibration model.

7. The video geolocation method based on tower monitoring according to claim 1, characterized in that: It also includes the steps to implement the function of finding videos by location: Select the area of interest in the remote sensing image and calculate the geographic coordinates of its center point; Find the reference image with the closest geographic coordinates through image retrieval; The geographic coordinates are inversely solved into image coordinates and projected into the real-time video frame.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the video geolocation method based on tower monitoring as described in any one of claims 1 to 7 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the video geo-positioning method based on tower monitoring as claimed in any one of claims 1 to 7 is implemented.

10. A computer program product comprising a computer program, characterized in that: When the computer program is executed by a processor, the video geo-positioning method based on tower monitoring as claimed in any one of claims 1 to 7 is implemented.

Images