Storage security inspection method and system

Abstract

The invention relates to a storage security inspection method and system, and belongs to the technical field of storage security. The method comprises the steps of obtaining a storage monitoring picture, constructing a storage security and protection data set through data expansion according to the storage monitoring picture, and constructing a storage security and protection model through an improved ReNet18 algorithm according to the storage security and protection data set; a three-dimensional patrol monitoring network is constructed, a high-definition storage monitoring video is obtained through the three-dimensional patrol monitoring network, a storage detection result is obtained through the storage security and protection model according to the high-definition storage monitoring video, and the storage detection result comprises no storage smoke and storage smoke abnormity. And sending a storage fire alarm instruction to the terminal according to the storage detection result, thereby realizing efficient storage inspection, automatically performing real-time analysis on the storage inspection content, accurately identifying and detecting the abnormal condition of storage smoke, timely generating storage fire alarm information, automatically responding to the abnormal condition, and ensuring the overall storage security and protection safety.

Description

Technical Field

[0001] The invention belongs to the technical field of warehouse security, and in particular relates to a warehouse security patrol method and system. Background Art

[0002] With the rapid development of social economy, the warehousing industry is becoming more and more prosperous, the scale of warehouses is constantly expanding, and the types and quantities of goods are increasing. As a result, there are more and more warehouse security issues, especially warehouse fire prevention. Due to the complex warehouse environment, high personnel mobility, and large fire safety hazards, a large number of security personnel are required to do a lot of repetitive patrols and monitoring work, which brings huge challenges to warehouse security. However, in the early stage of a warehouse fire, the flames are small and the smoke is thin. Manual patrols often fail to detect dangers in time, which can easily lead to huge losses and make it difficult to meet the growing demand for warehouse security. Moreover, most of the current research on warehouse fire smoke detection uses sensor-based and machine learning methods, which have the problems of slow detection time and low detection accuracy.

[0003] In view of the problems of low efficiency, high labor cost and high missed detection rate in the above-mentioned traditional warehouse security inspections, there is an urgent need to develop an efficient intelligent warehouse security inspection method. Through intelligent technology, the warehouse inspection content can be automatically analyzed in real time, and abnormal warehouse smoke conditions can be accurately identified and detected. Warehouse fire alarm information is generated in time to ensure the safety and stability of the warehouse, realize efficient warehouse inspections, liberate manpower, and ensure the overall warehouse security. Summary of the invention

[0004] In order to solve the above problems existing in the prior art, the present invention proposes a warehouse security patrol method and system.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] A warehouse security patrol method, comprising:

[0007] Acquire warehouse monitoring pictures, and construct a warehouse security data set through data expansion based on the warehouse monitoring pictures, wherein the warehouse monitoring pictures include warehouse smoke pictures and warehouse normal pictures;

[0008] Constructing a warehouse security model based on the warehouse security data set by improving the ReNet18 algorithm;

[0009] Establishing a three-dimensional patrol monitoring network, through which high-definition warehouse monitoring videos are acquired, wherein the three-dimensional patrol monitoring network includes mobile monitoring robots and drones;

[0010] A warehouse detection result is obtained through the warehouse security model according to the high-definition warehouse monitoring video, and the warehouse detection result includes a warehouse with no smoke and a warehouse with abnormal smoke. A warehouse fire alarm instruction is sent to the terminal according to the warehouse detection result.

[0011] Preferably, the step of acquiring a warehouse monitoring picture and constructing a warehouse security data set through data expansion according to the warehouse monitoring picture includes:

[0012] Collect public warehouse monitoring pictures from the Internet and screen the picture quality, while ensuring that all pictures are in the same format and adjusted to a uniform size;

[0013] The warehouse monitoring image data is expanded to obtain a warehouse security data set, and the data expansion includes rotation, scaling, color conversion, and noise addition.

[0014] Preferably, the rotation includes randomly rotating the warehouse monitoring picture at specific angles of 15°, 45°, 90°, and 270°;

[0015] The scaling includes randomly scaling the warehouse monitoring picture at a scaling ratio ranging from 0.75 to 1.5;

[0016] The color transformation includes randomly adjusting the brightness, contrast and saturation of the warehouse monitoring picture;

[0017] The adding of noise includes randomly adding Gaussian noise and salt and pepper noise to the warehouse monitoring picture.

[0018] Preferably, constructing a warehouse security model based on the warehouse security data set by improving the ReNet18 algorithm includes:

[0019] Performing image preprocessing on the warehouse monitoring images of the warehouse security data set, wherein the image preprocessing includes normalization processing and denoising processing;

[0020] According to the warehouse security data set, the warehouse security initial model is obtained by improving the ReNet18 network structure;

[0021] A warehouse security model is obtained by fine-tuning the warehouse security initial model according to the warehouse security data set.

[0022] Preferably, obtaining the warehouse security initial model by improving the ReNet18 network structure according to the warehouse security data set includes:

[0023] Initialize the ReNet18 network weights through transfer learning, where the transfer learning is to load the ReNet18 network pre-trained on the ImageNet dataset as the base model;

[0024] The improved ReNet18 network structure includes replacing convolution and introducing an attention mechanism. The replaced convolution is to replace the ordinary convolution in the ReNet18 network with GSConv, and the introduced attention mechanism is to introduce the ECA attention mechanism in the residual block of the ReNet18 network.

[0025] Preferably, the GSConv is used to obtain a first warehouse monitoring feature map by downsampling through ordinary convolution, obtain a second warehouse monitoring feature map by depthwise separable convolution, concatenate the first warehouse monitoring feature map and the second warehouse monitoring feature map to obtain an initial warehouse monitoring feature map, and obtain a warehouse monitoring feature map by shuffling the initial warehouse monitoring feature map;

[0026] The ECA attention mechanism is used to perform global average pooling on the warehouse monitoring feature map to obtain a warehouse monitoring pooling feature map, and to achieve local cross-channel interaction through one-dimensional convolution based on the warehouse monitoring pooling feature map to obtain the final warehouse monitoring feature map.

[0027] Preferably, the step of constructing a three-dimensional patrol monitoring network and obtaining high-definition warehouse monitoring videos through the three-dimensional patrol monitoring network includes:

[0028] A mobile monitoring robot patrols the warehouse along a preset path to obtain a high-definition first storage monitoring video, wherein the mobile monitoring robot is equipped with a high-definition camera;

[0029] A drone is used to cover the high altitude area of ​​the warehouse for aerial inspection to obtain a high-definition second surveillance video of the warehouse, wherein the drone is equipped with a high-definition camera;

[0030] A three-dimensional monitoring network is constructed by the UAV and the mobile monitoring robot, and the high-definition warehouse first monitoring video and the high-definition warehouse second monitoring video are integrated to obtain a high-definition warehouse monitoring video.

[0031] A warehouse security patrol system using the above-mentioned warehouse security patrol method includes a warehouse security smoke detection module, a three-dimensional patrol monitoring module, and a warehouse security alarm module;

[0032] The warehouse security smoke detection module is used to construct a warehouse security data set through data expansion, and to construct a warehouse security model based on the warehouse security data set by improving the ReNet18 algorithm;

[0033] The three-dimensional patrol monitoring module is used to build a three-dimensional monitoring network through drones and mobile monitoring robots, monitor the storage status in all directions, and obtain high-definition storage monitoring videos;

[0034] The warehouse security alarm module is used to obtain a warehouse detection result based on the high-definition warehouse monitoring video through the warehouse security model, and send a warehouse fire alarm instruction to the terminal according to the warehouse detection result.

[0035] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned warehouse security patrol method when executing the program.

[0036] A storage medium containing computer executable instructions, which are used to execute the above-mentioned warehouse security patrol method when executed by a computer processor.

[0037] The beneficial effects of the present invention are:

[0038] (1) A three-dimensional monitoring network is formed by setting up mobile monitoring robots and drone patrols. High-definition warehouse monitoring videos are collected based on the three-dimensional monitoring network. The high-definition warehouse monitoring videos are uploaded to the edge computing unit, intercepted into high-definition warehouse monitoring pictures and detected through the warehouse security model. Warehouse fire smoke is identified and warehouse fire alarm instructions are sent to the terminal. This solves the problem that fixed monitoring is difficult to identify small fires, realizes efficient warehouse patrols, liberates manpower, and ensures overall warehouse security.

[0039] (2) Through the collaborative work of mobile monitoring robots and drones, the mobile monitoring robots are responsible for ground inspections, while the drones are responsible for aerial monitoring. The two work together to ensure coverage without blind spots and achieve all-round, three-dimensional monitoring of the storage environment.

[0040] (3) By replacing the ordinary convolution in the ReNet18 network with GSConv, the model calculation amount and the number of parameters can be reduced, thereby reducing memory usage and computing resource consumption, and improving the real-time detection efficiency of the model while maintaining the detection accuracy; introducing the ECA attention mechanism in the residual block of the ReNet18 network can enhance the model's ability to extract key fire smoke features and further improve the model's detection performance.

[0041] (4) By initializing the ReNet18 network weights through transfer learning, the feature representation learned by the pre-trained model on a large-scale data set can be used to accelerate the convergence speed of model training; according to the warehouse security data set, the warehouse security model is obtained by fine-tuning the warehouse security initial model, which can optimize the warehouse fire smoke detection task and improve the efficiency and accuracy of warehouse fire smoke detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0043] Figure 1 The present invention is a flowchart of a warehouse security patrol method. DETAILED DESCRIPTION

[0044] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0045] See also Figure 1 , a warehouse security patrol method, comprising:

[0046] S1: Obtain warehouse monitoring pictures, and construct a warehouse security data set through data expansion based on the warehouse monitoring pictures, wherein the warehouse monitoring pictures include warehouse smoke pictures and warehouse normal pictures;

[0047] S2: Constructing a warehouse security model based on the warehouse security data set by improving the ReNet18 algorithm;

[0048] S3: constructing a three-dimensional patrol monitoring network, and obtaining high-definition warehouse monitoring videos through the three-dimensional patrol monitoring network, wherein the three-dimensional patrol monitoring network includes a mobile monitoring robot and a drone;

[0049] S4: Obtaining a warehouse detection result through the warehouse security model according to the high-definition warehouse monitoring video, wherein the warehouse detection result includes a warehouse without smoke and a warehouse with abnormal smoke, and sending a warehouse fire alarm instruction to the terminal according to the warehouse detection result.

[0050] In this embodiment, the acquisition of warehouse monitoring pictures and the construction of warehouse security data sets through data expansion according to the warehouse monitoring pictures are specifically implemented by the following steps:

[0051] S101: Collect public warehouse monitoring pictures from the Internet and screen the picture quality to remove blurry, dark or overexposed pictures. Ensure that all pictures are in the same format and adjust the pictures to a uniform size to facilitate subsequent processing.

[0052] It should be noted that the collected warehouse monitoring pictures include warehouse smoke pictures and normal warehouse pictures, and the proportions of the two pictures are comparable; at the same time, the warehouse monitoring pictures with uniform format and size after picture quality screening are labeled.

[0053] S102: The warehouse monitoring image data is expanded to obtain a warehouse security data set, and the data expansion includes rotation, scaling, color conversion, and noise addition.

[0054] Specifically, the warehouse monitoring picture is randomly rotated at specific angles of 15°, 45°, 90°, and 270° to simulate different monitoring viewing angles; the warehouse monitoring picture is randomly scaled at a scaling ratio ranging from 0.75 to 1.5 to cope with monitoring objects of different sizes; the brightness, contrast, and saturation of the warehouse monitoring picture are randomly adjusted to simulate different lighting conditions; Gaussian noise and salt and pepper noise are randomly added to the warehouse monitoring picture to simulate image quality problems that may occur in monitoring equipment.

[0055] In this embodiment, the warehouse security model is constructed by improving the ReNet18 algorithm according to the warehouse security data set through the following steps:

[0056] S201: performing image preprocessing on the warehouse monitoring images of the warehouse security data set, wherein the image preprocessing includes normalization processing and denoising processing;

[0057] S202: Obtaining an initial warehouse security model by improving the ReNet18 network structure according to the warehouse security data set;

[0058] S202-1: Initialize the ReNet18 network weights through transfer learning, wherein the transfer learning is to load the ReNet18 network pre-trained on the ImageNet dataset as a basic model;

[0059] S202-2: The improved ReNet18 network structure includes replacing convolution and introducing an attention mechanism, wherein the replacing convolution is to replace the ordinary convolution in the ReNet18 network with GSConv, and the introducing attention mechanism is to introduce an ECA attention mechanism in the residual block of the ReNet18 network;

[0060] S202-21: the GSConv is used to obtain a first warehouse monitoring feature map by downsampling through ordinary convolution, obtain a second warehouse monitoring feature map by depthwise separable convolution, concatenate the first warehouse monitoring feature map and the second warehouse monitoring feature map to obtain an initial warehouse monitoring feature map, and obtain a warehouse monitoring feature map by shuffling the initial warehouse monitoring feature map;

[0061] It should be noted that the shuffling operation causes the characteristic graphs of the channels corresponding to the first warehouse monitoring characteristic graph and the second warehouse monitoring characteristic graph to be arranged adjacent to each other.

[0062] S202-22: The ECA attention mechanism is used to perform global average pooling on the warehouse monitoring feature map to obtain a warehouse monitoring pooling feature map, and to achieve local cross-channel interaction through one-dimensional convolution based on the warehouse monitoring pooling feature map to obtain a warehouse monitoring final feature map;

[0063] It should be noted that the local cross-channel interaction is to generate the weights of each channel through the Sigmoid activation function, and then combine the warehouse monitoring feature map with the channel weights to obtain features with channel attention; the final feature map of warehouse monitoring is used to identify warehouse detection results when running the model.

[0064] S203: Obtaining a warehouse security model by fine-tuning the warehouse security initial model according to the warehouse security dataset.

[0065] Specifically, fine-tuning is performed on the warehouse security dataset to adapt to the smoke detection task in a warehouse environment, the output category of the fully connected layer of the pre-trained ReNet18 network is changed to two categories, and the warehouse security initial model is trained according to the warehouse security dataset;

[0066] The optimal model parameters are determined through cross-validation and hyperparameter tuning, and a warehouse security model is obtained based on the optimal model parameters.

[0067] Specifically, five-fold cross validation is used for model training, and the Adam optimizer is selected to improve training efficiency and quality, thereby optimizing the performance of the warehouse security model.

[0068] In this embodiment, the construction of a three-dimensional patrol monitoring network and the acquisition of high-definition warehouse monitoring video through the three-dimensional patrol monitoring network are specifically implemented by the following steps:

[0069] S301: patrolling the warehouse along a preset path by a mobile monitoring robot to obtain a high-definition first warehouse monitoring video, wherein the mobile monitoring robot is equipped with a high-definition camera;

[0070] S302: Using a drone to cover the high altitude area of ​​the warehouse for aerial inspection to obtain a high-definition second warehouse monitoring video, wherein the drone is equipped with a high-definition camera;

[0071] S303: A three-dimensional monitoring network is constructed by using the drone and the mobile monitoring robot, and the high-definition warehouse first monitoring video and the high-definition warehouse second monitoring video are integrated to obtain a high-definition warehouse monitoring video to monitor the warehouse status in all directions.

[0072] In this embodiment, obtaining a warehouse detection result through the warehouse security model according to the high-definition warehouse monitoring video, and sending a warehouse fire alarm instruction to the terminal according to the warehouse detection result includes:

[0073] S401: The high-definition warehouse monitoring video is uploaded to the edge computing unit and captured into a high-definition warehouse monitoring picture;

[0074] S402: Performing real-time detection and processing through the warehouse security model according to the high-definition warehouse monitoring picture to obtain the warehouse detection result;

[0075] When the warehouse detection result shows that the warehouse smoke is abnormal, an alarm instruction is immediately sent to the terminal to notify relevant personnel to take corresponding measures to ensure warehouse safety; at the same time, the detection results and alarm information are stored in the database for subsequent analysis and tracing.

[0076] A warehouse security patrol system includes a warehouse security smoke detection module, a three-dimensional patrol monitoring module, and a warehouse security alarm module;

[0077] The warehouse security smoke detection module is used to construct a warehouse security data set through data expansion, and to construct a warehouse security model based on the warehouse security data set by improving the ReNet18 algorithm;

[0078] The three-dimensional patrol monitoring module is used to build a three-dimensional monitoring network through drones and mobile monitoring robots, monitor the storage status in all directions, and obtain high-definition storage monitoring videos;

[0079] The warehouse security alarm module is used to obtain a warehouse detection result based on the high-definition warehouse monitoring video through the warehouse security model, and send a warehouse fire alarm instruction to the terminal according to the warehouse detection result.

[0080] The computer storage medium of the embodiment of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device.

[0081] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0082] The program code included on the computer readable medium can be transmitted with any appropriate medium, including but not limited to wireless, electric wire, optical cable, RF, etc., or any suitable combination of the above. The computer program code for performing the operation of the present invention can be written in one or more programming languages ​​or their combinations, and the programming language includes object-oriented programming languages-such as Java, Smalltalk, C++, and also includes conventional procedural programming languages-such as "C" language or similar programming languages. The program code can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on the remote computer, or completely on the remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0083] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A warehouse security patrol method, characterized in that: include: Obtaining warehouse monitoring pictures, and constructing a warehouse security data set through data expansion based on the warehouse monitoring pictures, wherein the warehouse monitoring pictures include warehouse smoke pictures and warehouse normal pictures; Constructing a warehouse security model based on the warehouse security data set by improving the ReNet18 algorithm; Establishing a three-dimensional patrol monitoring network, through which high-definition warehouse monitoring videos are obtained, wherein the three-dimensional patrol monitoring network includes mobile monitoring robots and drones; A warehouse detection result is obtained through the warehouse security model according to the high-definition warehouse monitoring video, and the warehouse detection result includes a warehouse with no smoke and a warehouse with abnormal smoke. A warehouse fire alarm instruction is sent to the terminal according to the warehouse detection result.

2. The warehouse security patrol method according to claim 1 is characterized in that: The obtaining of warehouse monitoring pictures and constructing a warehouse security data set through data expansion according to the warehouse monitoring pictures includes: Collect public warehouse monitoring pictures from the Internet and screen the picture quality, while ensuring that all pictures are in the same format and adjusted to a uniform size; The warehouse monitoring image data is expanded to obtain a warehouse security data set, and the data expansion includes rotation, scaling, color conversion, and noise addition.

3. The warehouse security patrol method according to claim 2 is characterized in that: The rotation includes randomly rotating the warehouse monitoring image at specific angles of 15°, 45°, 90°, and 270°; The scaling includes randomly scaling the warehouse monitoring picture at a scaling ratio ranging from 0.75 to 1.5; The color transformation includes randomly adjusting the brightness, contrast and saturation of the warehouse monitoring picture; The adding of noise includes randomly adding Gaussian noise and salt and pepper noise to the warehouse monitoring picture.

4. The warehouse security patrol method according to claim 1, characterized in that: The step of constructing a warehouse security model by improving the ReNet18 algorithm according to the warehouse security data set includes: Performing image preprocessing on the warehouse monitoring images of the warehouse security data set, wherein the image preprocessing includes normalization processing and denoising processing; According to the warehouse security data set, the warehouse security initial model is obtained by improving the ReNet18 network structure; A warehouse security model is obtained by fine-tuning the warehouse security initial model according to the warehouse security data set.

5. The warehouse security patrol method according to claim 4 is characterized in that: The warehouse security initial model obtained by improving the ReNet18 network structure according to the warehouse security data set includes: Initialize the ReNet18 network weights through transfer learning, where the transfer learning is to load the ReNet18 network pre-trained on the ImageNet dataset as the base model; The improved ReNet18 network structure includes replacing convolution and introducing an attention mechanism. The replaced convolution is to replace the ordinary convolution in the ReNet18 network with GSConv, and the introduced attention mechanism is to introduce the ECA attention mechanism in the residual block of the ReNet18 network.

6. The warehouse security patrol method according to claim 5 is characterized in that: The GSConv is used to obtain a first warehouse monitoring feature map by downsampling through ordinary convolution, obtain a second warehouse monitoring feature map by depthwise separable convolution, concatenate the first warehouse monitoring feature map and the second warehouse monitoring feature map to obtain an initial warehouse monitoring feature map, and obtain a warehouse monitoring feature map by shuffling the initial warehouse monitoring feature map; The ECA attention mechanism is used to perform global average pooling on the warehouse monitoring feature map to obtain a warehouse monitoring pooling feature map, and to achieve local cross-channel interaction through one-dimensional convolution based on the warehouse monitoring pooling feature map to obtain the final warehouse monitoring feature map.

7. The warehouse security patrol method according to claim 1, characterized in that: The constructing of a three-dimensional patrol monitoring network and obtaining high-definition warehouse monitoring videos through the three-dimensional patrol monitoring network include: A mobile monitoring robot patrols the warehouse along a preset path to obtain a high-definition first storage monitoring video, wherein the mobile monitoring robot is equipped with a high-definition camera; A drone is used to cover the high altitude area of ​​the warehouse for aerial inspection to obtain a high-definition second surveillance video of the warehouse, wherein the drone is equipped with a high-definition camera; A three-dimensional monitoring network is constructed by the UAV and the mobile monitoring robot, and the high-definition warehouse first monitoring video and the high-definition warehouse second monitoring video are integrated to obtain a high-definition warehouse monitoring video.

8. A warehouse security patrol system using the warehouse security patrol method of claim 1, comprising a warehouse security smoke detection module, a three-dimensional patrol monitoring module, and a warehouse security alarm module, characterized in that: The warehouse security smoke detection module is used to construct a warehouse security data set through data expansion, and to construct a warehouse security model based on the warehouse security data set by improving the ReNet18 algorithm; The three-dimensional patrol monitoring module is used to build a three-dimensional monitoring network through drones and mobile monitoring robots, monitor the storage status in all directions, and obtain high-definition storage monitoring videos; The warehouse security alarm module is used to obtain a warehouse detection result based on the high-definition warehouse monitoring video through the warehouse security model, and send a warehouse fire alarm instruction to the terminal according to the warehouse detection result.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the warehouse security patrol method as described in any one of claims 1-7 is implemented.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions, when executed by a computer processor, are used to execute the warehouse security patrol method as described in any one of claims 1-7.

Images