Method and device for recognizing abnormal object in elevator based on visual neural network

Abstract

The invention discloses a method and device for recognizing an abnormal object in an elevator based on an optic neural network, and the method comprises the steps: collecting a first video and a second video through a first camera and a second camera, inputting the first video and the second video into an electric vehicle recognition neural network to obtain a first score, and judging whether the first score is greater than a first score threshold value or not; if yes, it is determined that an electric vehicle exists, if not, when the first score is larger than a second score threshold value, first weights of all objects in the elevator car and second weights of predicted passengers are obtained, and whether the electric vehicle exists or not is judged after the first score is corrected through the first weights and the second weights to obtain a second score; the shielded electric vehicle is recognized through videos collected by the cameras at different angles, whether the electric vehicle exists or not is judged after the score is corrected through the weight, the accuracy and reliability of recognizing the electric vehicle are higher, the electric vehicle can be accurately and reliably reminded to enter the elevator, potential safety hazards caused when the electric vehicle enters the elevator are eliminated, and the safety of the elevator is improved. And the safety performance of the elevator is improved.

Description

Technical Field

[0001] The present invention relates to the field of elevator safety technology, and in particular to a method and device for identifying abnormal objects in an elevator based on a visual neural network. Background Art

[0002] Elevators are usually installed in high-rise residential buildings to facilitate residents to go up and down floors and improve travel efficiency. However, with the convenience of electric vehicles, residents often push electric vehicles into elevators, which poses safety hazards such as battery fires and explosions.

[0003] In the prior art, after the camera in the elevator collects the image inside the car, a visual neural network is used to identify from the image whether there is an electric car in the car. If so, a voice alarm is issued and the car door and hall door are controlled to open, prompting the residents to push the electric car out of the car.

[0004] However, electric vehicles in the elevator cabin are usually blocked by passengers in the elevator, which reduces the recognition accuracy of electric vehicles by the visual neural network and often misidentifies electric vehicles as ordinary bicycles. Summary of the invention

[0005] The present invention provides a method and device for identifying abnormal objects in an elevator based on a visual neural network, so as to solve the problem that an electric bicycle cannot be identified or is misidentified due to being blocked by a passenger.

[0006] In a first aspect, the present invention provides a method for identifying abnormal objects in an elevator based on a visual neural network, which is used to identify electric vehicles in an elevator, comprising:

[0007] When the elevator door is opened, a first camera in the elevator car is controlled to collect a first video and a second camera is controlled to collect a second video, wherein the first camera and the second camera are cameras at different positions in the elevator car;

[0008] Inputting the first video and the second video into an electric vehicle recognition neural network to obtain a first score of the electric vehicle in the elevator car;

[0009] Determining whether the first score is greater than a preset first score threshold;

[0010] If so, determining that there is an electric vehicle in the elevator car and generating an alarm message;

[0011] If not, when the first score is greater than a preset second score threshold, starting a weighing sensor to obtain a first weight of all objects in the elevator car, and the second score threshold is less than the first score threshold;

[0012] Inputting at least one of the first video and the second video into a weight prediction model to obtain a second weight of the passenger in the elevator car;

[0013] Correcting the first score using the first weight and the second weight to obtain a second score;

[0014] Determining whether the second score is greater than the first score threshold;

[0015] If so, executing the step of determining that there is an electric vehicle in the elevator car and generating an alarm message;

[0016] If not, it is determined that there is no electric vehicle in the elevator car.

[0017] Optionally, the electric vehicle recognition neural network includes a first feature extraction network, a second feature extraction network, a feature fusion network, a first detection head, and a second detection head. The first video and the second video are input into the electric vehicle recognition neural network to obtain a first score indicating the presence of an electric vehicle in the elevator car, including:

[0018] Inputting the first video into a first feature extraction network, and extracting a first feature map from the first video in the first feature extraction network;

[0019] Inputting the second video into a second feature extraction network, and extracting a second feature map from the second video in the second feature extraction network;

[0020] fusing the first feature map and the second feature map in the feature fusion network to obtain a fused feature;

[0021] The first detection head and the second detection head respectively identify the fusion features to obtain a first initial score of the electric vehicle in the first video and a second initial score of the electric vehicle in the second video;

[0022] An average of the first initial score and the second initial score is calculated as a first score for the presence of an electric vehicle in the elevator car.

[0023] Optionally, the electric vehicle identification neural network is trained in the following manner:

[0024] A first image and a second image of an elevator car taken from different angles are obtained as training images, wherein the first image is annotated with a third fraction of electric vehicles and a first occlusion rate of the electric vehicles, and the second image is annotated with a fourth fraction of electric vehicles and a second occlusion rate of the electric vehicles;

[0025] Inputting the first image and the second image into an electric vehicle recognition neural network to obtain a fifth score indicating the presence of an electric vehicle in the first image and a sixth score indicating the presence of an electric vehicle in the second image;

[0026] Calculating the loss rate using the third score, the fourth score, the fifth score, the sixth score, the first occlusion rate, and the second occlusion rate;

[0027] Determine whether the preset training conditions are met;

[0028] If so, determining that the electric vehicle identification neural network has completed training;

[0029] If not, adjust the network parameters of the electric vehicle identification neural network according to the loss rate, and return to the step of inputting the first image and the second image into the electric vehicle identification neural network.

[0030] Optionally, the loss rate is calculated using the third score, the fourth score, the fifth score and the sixth score, including:

[0031] calculating a first loss rate by using the third score and the fifth score, and calculating a second loss rate by using the fourth score and the sixth score;

[0032] Calculating a first product of the first loss rate and the first occlusion rate, and calculating a second product of the second loss rate and the second occlusion rate;

[0033] The loss rate is obtained by calculating the sum of the first product and the second product.

[0034] Optionally, the weight prediction model is trained in the following manner:

[0035] Acquire a sample image set, wherein the sample images in the sample image set are images when there are passengers in the elevator car, and each sample image is marked with a third weight;

[0036] Inputting the sample image into a weight prediction model to obtain a fourth weight of the passenger;

[0037] calculating a loss rate using the third weight and the fourth weight;

[0038] Determine whether the preset training conditions are met;

[0039] If so, determining that the weight prediction model has completed training;

[0040] If not, the model parameters of the weight prediction model are adjusted according to the loss rate, and the process returns to the step of inputting the sample image into the weight prediction model to obtain the second weight of the passenger.

[0041] Optionally, correcting the first score by using the first weight and the second weight to obtain a second score includes:

[0042] calculating a correction factor based on the first weight and the second weight;

[0043] The product of the correction coefficient and the first fraction is calculated to obtain a second fraction.

[0044] Optionally, calculating a correction factor based on the first weight and the second weight includes:

[0045] calculating a difference between the first weight and the second weight;

[0046] Determine whether the difference is greater than a value of 0;

[0047] If not, executing the step of determining that there is no electric vehicle in the elevator car;

[0048] If so, when the difference is greater than a preset weight threshold, calculating a ratio of the difference to the preset weight;

[0049] The correction factor is determined from the ratio in a preconfigured ratio-factor table.

[0050] In a second aspect, the present invention provides a device for identifying abnormal objects in an elevator based on a visual neural network, which is used to identify electric vehicles in an elevator, comprising:

[0051] A video acquisition module, used for controlling a first camera in the elevator car to acquire a first video and controlling a second camera to acquire a second video when the elevator door is opened, wherein the first camera and the second camera are cameras at different positions in the elevator car;

[0052] A first score prediction module, configured to input the first video and the second video into an electric vehicle recognition neural network to obtain a first score of the electric vehicle in the elevator car;

[0053] A first score determination module, used to determine whether the first score is greater than a preset first score threshold, and if so, execute a first determination module, and if not, execute a first weight acquisition module;

[0054] A first determination module, used to determine the presence of an electric vehicle in the elevator car and generate an alarm message;

[0055] a first weight acquisition module, configured to start a weighing sensor to acquire a first weight of all objects in the elevator car when the first score is greater than a preset second score threshold, and the second score threshold is less than the first score threshold;

[0056] A second weight prediction module, configured to input at least one of the first video and the second video into a weight prediction model to obtain a second weight of the passenger in the elevator car;

[0057] A second score correction module, configured to correct the first score by using the first weight and the second weight to obtain a second score;

[0058] A second score determination module, used to determine whether the second score is greater than the first score threshold, and if so, execute the first determination module, and if not, execute the second determination module;

[0059] The second determining module is used to determine whether there is no electric vehicle in the elevator car.

[0060] In a third aspect, the present invention provides an electronic device, the electronic device comprising:

[0061] at least one processor; and

[0062] a memory communicatively connected to the at least one processor; wherein,

[0063] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for identifying abnormal objects in an elevator based on a visual neural network as described in any one of the first aspects of the present invention.

[0064] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for identifying abnormal objects in an elevator based on a visual neural network as described in any one of the first aspects of the present invention when executed.

[0065] In an embodiment of the present invention, a first camera and a second camera at different positions in an elevator car respectively collect a first video and a second video, input the first video and the second video into an electric vehicle recognition neural network to obtain a first score of the electric vehicle in the elevator car, and judge whether the first score is greater than a preset first score threshold. If so, it is determined that there is an electric vehicle in the elevator car and an alarm message is generated. If not, when the first score is greater than a preset second score threshold, a weighing sensor is started to obtain a first weight of all objects in the elevator car, wherein the second score threshold is less than the first score threshold, at least one of the first video and the second video is input into a weight prediction model to obtain a second weight of passengers in the elevator car, the first score is corrected by the first weight and the second weight to obtain a second score, and it is judged whether the second score is greater than the first score threshold. If so, it is determined that there is an electric vehicle in the elevator car and an alarm message is generated. If not, it is determined that there is no electric vehicle in the elevator car. It realizes the identification of electric vehicles by inputting videos captured by cameras at different angles into the electric vehicle recognition neural network, avoiding the problem of failure to identify or misidentification of electric vehicles when they are blocked, improving the accuracy of electric vehicle recognition, and further correcting the score by weight when the score is low. For example, when an electric vehicle is identified as an ordinary bicycle and the score is low, the weight correction can be used to determine whether an electric vehicle exists by the corrected second score. The accuracy and reliability of identifying electric vehicles are higher, and electric vehicles can be accurately and reliably reminded to enter the elevator, eliminating the safety hazard of electric vehicles entering the elevator and improving the safety performance of the elevator.

[0066] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0068] Figure 1 is a flow chart of a method for identifying abnormal objects in an elevator based on a visual neural network provided in Embodiment 1 of the present invention;

[0069] Figure 2 This is a schematic diagram of the installation of the camera in the elevator car;

[0070] Figure 3 is a flow chart of a method for identifying abnormal objects in an elevator based on a visual neural network provided in Embodiment 2 of the present invention;

[0071] Figure 4 is a schematic diagram of the structure of an electric vehicle identification neural network in an embodiment of the present invention;

[0072] Figure 5 It is a structural schematic diagram of a device for identifying abnormal objects in an elevator based on a visual neural network provided in Embodiment 3 of the present invention;

[0073] Figure 6 It is a schematic diagram of the structure of an electronic device provided in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0074] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0075] Embodiment 1

[0076] Figure 1 This is a flowchart of a method for identifying abnormal objects in an elevator based on a visual neural network provided in the first embodiment of the present invention. This embodiment can be applied to identifying the presence of electric vehicles in an elevator. The method can be performed by a device for identifying abnormal objects in an elevator based on a visual neural network. The device for identifying abnormal objects in an elevator based on a visual neural network can be implemented in the form of hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method for identifying abnormal objects in an elevator based on a visual neural network includes:

[0077] S101. When an elevator door is opened, control a first camera in the elevator car to collect a first video and control a second camera to collect a second video, wherein the first camera and the second camera are cameras at different positions in the elevator car.

[0078] The embodiment of the present invention is used to detect whether there is an electric vehicle in an elevator car, wherein the electric vehicle may refer to an electric bicycle or an electric motorcycle. Since passengers usually get on and off the elevator only when the elevator door is open, the camera can be triggered to collect video in the elevator car when the elevator door is opened.

[0079] In one embodiment, a first camera and a second camera may be provided in the elevator car, and the first camera and the second camera are installed at different positions in the elevator car. Preferably, Figure 2As shown in the overhead view of the elevator car, the first camera and the second camera can be set at 90°, that is, the first camera and the second camera are installed on two adjacent car walls of the elevator car, so that the video of the electric bicycle can be captured at the maximum angle when the electric bicycle enters the elevator car in any posture.

[0080] When the car door is opened, the first camera can be used to capture the first video and the second camera can be controlled to capture the second video.

[0081] S102, inputting the first video and the second video into an electric vehicle recognition neural network to obtain a first score of the electric vehicle in the elevator car.

[0082] This embodiment can pre-train an electric vehicle recognition neural network, which has dual-stream input. After inputting the first video and the second video in the elevator car captured at different angles, the electric vehicle recognition neural network identifies whether there is an electric vehicle in the elevator car from two angles. This can avoid identifying the electric vehicle from the video captured from another direction when the electric vehicle is blocked by passengers in one direction, and realizes the fusion of video features from two directions to obtain features of the electric vehicle at multiple angles, thereby improving the accuracy of recognition when the electric vehicle is blocked. The higher the first score, the higher the reliability of identifying the electric vehicle.

[0083] S103: Determine whether the first score is greater than a preset first score threshold.

[0084] Among them, the first score threshold can be a score indicating a higher reliability of identifying an electric vehicle. For example, taking 100 as a full score, the first score threshold can be 90 points. The first score can be compared with the first score threshold. If the first score is greater than the first score threshold, S104 can be executed. If the first score is less than the first score threshold, S105 can be executed.

[0085] S104: Determine that there is an electric vehicle in the elevator car and generate an alarm message.

[0086] If the first score is greater than the first score threshold, it is determined that there is an electric vehicle in the elevator car. For example, when a passenger pushes an electric vehicle into the elevator car, and there are only a small number of passengers in the elevator car that do not block the electric vehicle, the electric vehicle recognition neural network can accurately and reliably identify the electric vehicle through the first video and the second video. The first score is greater than the first score threshold, and the reliability of identifying the electric vehicle is high. The presence of the electric vehicle in the elevator car can be directly determined, and an alarm message can be generated to trigger the voice device in the elevator car to broadcast a prompt sound, control the elevator car door to open, and control the elevator car to stop running, so as to prompt the passenger to push the electric vehicle out of the elevator car.

[0087] S105. When the first score is greater than a preset second score threshold, start the weighing sensor to obtain the first weight of all objects in the elevator car, and the second score threshold is less than the first score threshold.

[0088] Among them, the second score threshold may be a score less than the first score threshold, and the second score threshold may be a score indicating that there is a high reliability that there is no electric vehicle in the elevator car. Exemplarily, if the first score threshold is 90 points, the second score threshold may be 60 points. If the first score is less than the first score threshold and greater than the second score threshold, it indicates that there is a suspected electric vehicle in the elevator car and further confirmation is required. The total weight of all objects in the elevator car can be obtained as the first weight through the weighing sensor on the elevator. Exemplarily, the total weight can be collected by the weight sensor of the elevator car and then the weight of the car can be removed to obtain the first weight. If the first score is less than the second score threshold, it indicates that there is a high reliability that there is no electric vehicle in the elevator car, and S109 can be executed to determine that there is no electric vehicle in the elevator car.

[0089] S106. Input at least one of the first video and the second video into a weight prediction model to obtain a second weight of the passengers in the elevator car.

[0090] This embodiment may pre-train a weight prediction model, which predicts a second weight through a video in the elevator car. The second weight may be the weight of all objects (passengers and items) in the elevator car except electric vehicles or suspected electric vehicles.

[0091] S107. Correct the first score using the first weight and the second weight to obtain a second score.

[0092] In one embodiment, the difference between the first weight and the second weight can be calculated, and the difference represents the weight of the object suspected to be an electric vehicle in the elevator car. In view of the fact that electric vehicles are often misidentified as ordinary bicycles, and the weight of electric vehicles is usually 3-5 times that of ordinary bicycles, if the difference is less than 0, S109 can be executed to determine that there is no electric vehicle in the elevator car. If the difference is greater than 0, the ratio of the difference to the preset weight (the weight of most electric vehicles) can be calculated, and then the correction coefficient can be determined in the pre-configured ratio-coefficient table, and the product of the correction coefficient and the first score is calculated to obtain the second score.

[0093] S108: Determine whether the second score is greater than the first score threshold.

[0094] Specifically, after calculating the corrected second score, it can be determined whether the second score is greater than the first score threshold. If so, S104 is executed to determine that there is an electric car in the elevator car and generate an alarm message. If the second score is less than the first score threshold, S109 can be executed.

[0095] S109: Determine whether there is no electric vehicle in the elevator car.

[0096] If the corrected second score is less than the first score threshold, it is determined that there is no electric vehicle in the elevator car, and the elevator door may be allowed to close and the elevator car may be allowed to run.

[0097] In an embodiment of the present invention, a first camera and a second camera at different positions in an elevator car respectively collect a first video and a second video, input the first video and the second video into an electric vehicle recognition neural network to obtain a first score of the electric vehicle in the elevator car, and judge whether the first score is greater than a preset first score threshold. If so, it is determined that there is an electric vehicle in the elevator car and an alarm message is generated. If not, when the first score is greater than a preset second score threshold, a weighing sensor is started to obtain a first weight of all objects in the elevator car, wherein the second score threshold is less than the first score threshold, at least one of the first video and the second video is input into a weight prediction model to obtain a second weight of passengers in the elevator car, the first score is corrected by the first weight and the second weight to obtain a second score, and it is judged whether the second score is greater than the first score threshold. If so, it is determined that there is an electric vehicle in the elevator car and an alarm message is generated. If not, it is determined that there is no electric vehicle in the elevator car. It realizes the identification of electric vehicles by inputting videos captured by cameras at different angles into the electric vehicle recognition neural network, avoiding the problem of failure to identify or misidentification of electric vehicles when they are blocked, improving the accuracy of electric vehicle recognition, and further correcting the score by weight when the score is low. For example, when an electric vehicle is identified as an ordinary bicycle and the score is low, the weight correction can be used to determine whether an electric vehicle exists by the corrected second score. The accuracy and reliability of identifying electric vehicles are higher, and electric vehicles can be accurately and reliably reminded to enter the elevator, eliminating the safety hazard of electric vehicles entering the elevator and improving the safety performance of the elevator.

[0098] Embodiment 2

[0099] Figure 3 A flowchart of a method for identifying abnormal objects in an elevator based on a visual neural network is provided in the second embodiment of the present invention. The embodiment of the present invention is optimized on the basis of the above-mentioned first embodiment, such as Figure 3 As shown, the method for identifying abnormal objects in an elevator based on a visual neural network includes:

[0100] S301. When the elevator door is opened, control the first camera in the elevator car to collect a first video and control the second camera to collect a second video, wherein the first camera and the second camera are cameras at different positions in the elevator car.

[0101] like Figure 2As shown in the overhead view of the elevator car, the first camera and the second camera can be set at 90°, that is, the first camera and the second camera are installed on two adjacent car walls of the elevator car. When the elevator door is opened, the first camera can be controlled to capture the first video and the second camera can be controlled to capture the second video.

[0102] S302, inputting the first video into a first feature extraction network, extracting a first feature map for the first video in the first feature extraction network, inputting the second video into a second feature extraction network, and extracting a second feature map for the second video in the second feature extraction network.

[0103] like Figure 4 The figure shows a schematic diagram of the structure of an electric vehicle identification neural network according to an embodiment of the present invention. The electric vehicle identification neural network includes a first feature extraction network, a second feature extraction network, a feature fusion network, a first detection head and a second detection head. The electric vehicle identification neural network can be trained by the following steps:

[0104] S1. Obtain a first image and a second image of an elevator car taken from different angles as training images, wherein the first image is annotated with a third score of electric vehicles and a first occlusion rate of electric vehicles, and the second image is annotated with a fourth score of electric vehicles and a second occlusion rate of electric vehicles;

[0105] S2, inputting the first image and the second image into an electric vehicle recognition neural network to obtain a fifth score of the presence of an electric vehicle in the first image and a sixth score of the presence of an electric vehicle in the second image;

[0106] S3, calculating the loss rate using the third score, the fourth score, the fifth score, the sixth score, the first occlusion rate and the second occlusion rate;

[0107] S4: Determine whether the preset training conditions are met. If yes, execute S5; if no, execute S6

[0108] S5, determining that the electric vehicle recognition neural network has completed training;

[0109] S6. Adjust the network parameters of the electric vehicle identification neural network according to the loss rate, and return to S2.

[0110] Exemplarily, the first image is input into the first feature extraction network to extract the first feature map, the second image is input into the second feature extraction network to extract the second feature map, the first feature map and the second feature map are input into the feature fusion network for feature fusion to obtain fused features, and the fused features are input into the first detection head and the second detection head to obtain the fifth score and the sixth score respectively.

[0111] When calculating the loss rate, the first loss rate (such as mean square error loss) can be calculated by the third score and the fifth score, and the second loss rate (such as mean square error loss) can be calculated by the fourth score and the sixth score, and the first product of the first loss rate and the first occlusion rate can be calculated, and the second product of the second loss rate and the second occlusion rate can be calculated, and the sum of the first product and the second product can be calculated to obtain the loss rate. The loss rate is calculated by using the occlusion rate as a weight, and the network parameters are adjusted by the loss rate, so that the electric vehicle recognition neural network can learn the ability to recognize obscured electric vehicles.

[0112] After the electric vehicle identification neural network is trained and deployed, the first video can be input into the first feature extraction network to extract a first feature map for the first video in the first feature extraction network, and the second video can be input into the second feature extraction network to extract a second feature map for the second video in the second feature extraction network.

[0113] S303: Fusing the first feature map and the second feature map in a feature fusion network to obtain a fusion feature.

[0114] Specifically, the first occlusion rate may be used as the weight of the first feature map, the second occlusion rate may be used as the weight of the second feature map, and the first feature map and the second feature map may be fused by the weights to obtain a fused feature.

[0115] S304, identifying the fused features at the first detection head and the second detection head respectively, to obtain a first initial score of the presence of electric vehicles in the first video and a second initial score of the presence of electric vehicles in the second video.

[0116] Specifically, the fused features may be input into the first detection head to obtain a first initial score indicating the presence of an electric vehicle in the first video, and the fused features may be input into the second detection head to obtain a second initial score indicating the presence of an electric vehicle in the second video.

[0117] S305. Calculate the average of the first initial score and the second initial score as a first score for the presence of the electric vehicle in the elevator car.

[0118] Specifically, after obtaining the first initial score and the second initial score, the arithmetic mean of the first initial score and the second initial score may be calculated as the first score.

[0119] In another embodiment, the weight of the first camera can be configured as the weight of the first initial score, and the weight of the second camera can be configured as the weight of the second initial score. The weighted average is calculated by the weight, the first initial score and the second initial score as the first score. For example, if the first camera is set on one of the left and right side walls of the elevator car, the first camera can capture the side of the electric car after the electric car enters the car (the electric car is usually located in the elevator car according to the posture when entering, so a larger weight can be set). The weight of the first camera can be 0.6, and the second camera is set on one of the front and rear side walls of the elevator car, and the weight of the second camera can be 0.4.

[0120] S306: Determine whether the first score is greater than a preset first score threshold.

[0121] Specifically, the first score threshold may be a score indicating a higher reliability of identifying an electric vehicle. The first score may be compared with the first score threshold. If the first score is greater than the first score threshold, S307 may be executed. If the first score is less than the first score threshold, S308 may be executed.

[0122] S307: Determine that there is an electric vehicle in the elevator car and generate an alarm message.

[0123] If the first score is greater than the first score threshold, it can be determined that there is an electric vehicle in the elevator car, and an alarm message can be generated to trigger the voice device in the elevator car to broadcast a prompt sound, control the elevator car door to open, and control the elevator car to stop running, so as to prompt the passengers to push the electric vehicle out of the elevator car.

[0124] S308. When the first score is greater than a preset second score threshold, start the weighing sensor to obtain the first weight of all objects in the elevator car, and the second score threshold is less than the first score threshold.

[0125] The second score threshold may be a score less than the first score threshold, and the second score threshold may be a score with high reliability indicating that there is no electric vehicle in the elevator car. If the first score is less than the first score threshold and greater than the second score threshold, it indicates that there is suspected electric vehicle in the elevator car, and further confirmation is required. The total weight of all objects in the elevator car can be obtained as the first weight and the real weight through the weighing sensor on the elevator.

[0126] S309, input at least one of the first video and the second video into a weight prediction model to obtain a second weight of the passengers in the elevator car.

[0127] In this embodiment, the second weight may be an estimated weight of objects other than the suspected electric vehicle in the elevator car, which may be predicted by a weight prediction model, wherein the weight prediction model may be trained by the following steps:

[0128] S10, obtaining a sample image set, wherein the sample images in the sample image set are images when there are passengers in the elevator car, and each sample image is marked with a third weight;

[0129] S11, inputting the sample image into the weight prediction model to obtain the fourth weight of the passenger;

[0130] S12, calculating the loss rate using the third weight and the fourth weight;

[0131] S13, determining whether preset training conditions are met;

[0132] S14, determining that the weight prediction model has completed training;

[0133] S15. Adjust the model parameters of the weight prediction model according to the loss rate, and return to S11.

[0134] The specific training process may refer to the supervised training method of the neural network model in the prior art, and this embodiment does not limit the training process.

[0135] After the weight prediction model is trained and deployed, at least one of the first video and the second video can be input into the weight prediction model, and a frame of image is extracted from the input video and input into the weight prediction model to predict the estimated weight of objects (such as passengers) in the car other than the suspected electric vehicle as the second weight.

[0136] S310, calculating a correction coefficient based on the first weight and the second weight.

[0137] In one embodiment, the first weight is the actual weight, and the second weight is the estimated weight of objects other than the object suspected to be an electric vehicle. The difference between the first weight and the second weight can be calculated, and the difference represents the gravity of the object suspected to be an electric vehicle. It can be determined whether the difference is greater than the value 0. If not, S313 can be executed to determine whether there is no electric vehicle in the elevator car. If so, the ratio of the difference to the preset weight is calculated when the difference is greater than the preset weight threshold, and the sum of the value 1 and the ratio is calculated. The correction coefficient is determined in a pre-configured ratio-coefficient table through the ratio, wherein the preset weight can be the weight of most electric vehicles.

[0138] For example, if the first weight is 640KG, the second weight is 600KG, and the difference is 40KG, which is greater than the weight threshold of 20KG, the ratio of the difference 40KG to the preset weight (50KG) can be calculated as 0.8, and then the correction coefficient corresponding to 0.8 is determined to be 1.2 in the ratio-coefficient table.

[0139] If the first weight is 640KG and the second weight is 630KG, the difference is 10KG, which is less than the weight threshold of 20KG. That is, the difference between the actual first weight of 60KG and the second weight of 630KG after removing objects suspected of being electric vehicles (such as ordinary bicycles) is small. It is possible that it is indeed an ordinary bicycle, and it can be determined that there is no electric vehicle in the elevator car.

[0140] S311. Calculate the product of the correction coefficient and the first score to obtain a second score.

[0141] After the calibration coefficient is obtained, the product of the calibration coefficient and the first fraction may be calculated to obtain the second fraction.

[0142] S312: Determine whether the second score is greater than the first score threshold.

[0143] Specifically, after calculating the corrected second score, it can be determined whether the second score is greater than the first score threshold. If so, S307 is executed to determine that there is an electric car in the elevator car and generate an alarm message. If the second score is less than the first score threshold, S313 can be executed.

[0144] S313: Determine whether there is no electric vehicle in the elevator car.

[0145] If the corrected second score is less than the first score threshold, it is determined that there is no electric vehicle in the elevator car, and the elevator door may be allowed to close and the elevator car may be allowed to run.

[0146] In this embodiment, the videos captured by the first camera and the second camera in the car are respectively input into the first feature extraction network and the second feature map extraction network of the electric vehicle recognition neural network to obtain the first feature map and the second feature map, and after fusion through the feature fusion network, the first initial score is output through the first detection head and the second initial score is output through the second detection head, and the average of the first initial score and the second initial score is calculated as the first score, and it is judged whether the first score is greater than the preset first score threshold. If so, it is determined that there is an electric vehicle in the elevator car and an alarm message is generated. If not, when the first score is greater than the preset second score threshold, the weighing sensor is started to obtain the first weight of all objects in the elevator car, wherein the second score threshold is less than the first score threshold, at least one of the first video and the second video is input into the weight prediction model to obtain the second weight of the passengers in the elevator car, and the first score is corrected by the first weight and the second weight to obtain the second score, and it is judged whether the second score is greater than the first score threshold. If so, it is determined that there is an electric vehicle in the elevator car and an alarm message is generated. If not, it is determined that there is no electric vehicle in the elevator car. It realizes the identification of electric vehicles by inputting videos captured by cameras at different angles into the electric vehicle recognition neural network, avoiding the problem of failure to identify or misidentification of electric vehicles when they are blocked, improving the accuracy of electric vehicle recognition, and further correcting the score by weight when the score is low. For example, when an electric vehicle is identified as an ordinary bicycle and the score is low, the weight correction can be used to determine whether an electric vehicle exists by the corrected second score. The accuracy and reliability of identifying electric vehicles are higher, and electric vehicles can be accurately and reliably reminded to enter the elevator, eliminating the safety hazard of electric vehicles entering the elevator and improving the safety performance of the elevator.

[0147] Embodiment 3

[0148] Figure 5 This is a schematic diagram of the structure of a device for identifying abnormal objects in an elevator based on a visual neural network provided in the third embodiment of the present invention. Figure 5 As shown, the device for identifying abnormal objects in an elevator based on a visual neural network includes:

[0149] The video acquisition module 501 is used to control a first camera in the elevator car to acquire a first video and control a second camera to acquire a second video when the elevator door is opened, wherein the first camera and the second camera are cameras at different positions in the elevator car;

[0150] A first score prediction module 502, configured to input the first video and the second video into an electric vehicle recognition neural network to obtain a first score of the electric vehicle in the elevator car;

[0151] A first score judging module 503 is used to judge whether the first score is greater than a preset first score threshold, and if so, execute a first determining module 504, and if not, execute a first weight obtaining module 505;

[0152] A first determination module 504 is used to determine that there is an electric vehicle in the elevator car and generate an alarm message;

[0153] A first weight acquisition module 505, configured to start a weighing sensor to acquire a first weight of all objects in the elevator car when the first score is greater than a preset second score threshold, and the second score threshold is less than the first score threshold;

[0154] A second weight prediction module 506, configured to input at least one of the first video and the second video into a weight prediction model to obtain a second weight of the passenger in the elevator car;

[0155] A second score correction module 507, configured to correct the first score by using the first weight and the second weight to obtain a second score;

[0156] A second score determination module 508, used to determine whether the second score is greater than the first score threshold, if so, execute the first determination module 504, if not, execute the second determination module 509;

[0157] The second determining module 509 is used to determine whether there is no electric vehicle in the elevator car.

[0158] Optionally, the electric vehicle recognition neural network includes a first feature extraction network, a second feature extraction network, a feature fusion network, a first detection head and a second detection head, and the first score prediction module 502 includes:

[0159] A first feature map extraction unit, configured to input a first video into a first feature extraction network, and extract a first feature map from the first video in the first feature extraction network;

[0160] A second feature map extraction unit, used for inputting the second video into a second feature extraction network, and extracting a second feature map from the second video in the second feature extraction network;

[0161] A feature fusion unit, used for fusing the first feature map and the second feature map in the feature fusion network to obtain a fusion feature;

[0162] A score prediction unit is used to identify the fusion feature at the first detection head and the second detection head respectively, to obtain a first initial score of the electric vehicle in the first video and a second initial score of the electric vehicle in the second video;

[0163] The score calculation unit is used to calculate the average of the first initial score and the second initial score as a first score for the presence of the electric vehicle in the elevator car.

[0164] Optionally, the electric vehicle recognition neural network is trained by the following modules:

[0165] a training image acquisition module, used to acquire a first image and a second image of the elevator car taken from different angles as training images, wherein the first image is annotated with a third fraction of electric vehicles and a first occlusion rate of the electric vehicles, and the second image is annotated with a fourth fraction of electric vehicles and a second occlusion rate of the electric vehicles;

[0166] a score prediction module, configured to input the first image and the second image into an electric vehicle recognition neural network to obtain a fifth score indicating the presence of an electric vehicle in the first image and a sixth score indicating the presence of an electric vehicle in the second image;

[0167] An electric vehicle identification loss rate calculation module, configured to calculate the loss rate using the third score, the fourth score, the fifth score, the sixth score, the first occlusion rate, and the second occlusion rate;

[0168] The electric vehicle identification neural network training condition judgment module is used to judge whether the preset training conditions are met. If so, the electric vehicle identification neural network training completion determination module is executed; if not, the electric vehicle identification neural network adjustment module is executed;

[0169] An electric vehicle identification neural network training completion determination module is used to determine that the electric vehicle identification neural network training is completed;

[0170] The electric vehicle identification neural network adjustment module is used to adjust the network parameters of the electric vehicle identification neural network according to the loss rate and return to the score prediction module.

[0171] Optionally, the electric vehicle identification loss rate calculation module includes:

[0172] a first loss rate calculation unit, configured to calculate a first loss rate by using the third score and the fifth score, and to calculate a second loss rate by using the fourth score and the sixth score;

[0173] a second loss rate calculation unit, configured to calculate a first product of the first loss rate and the first shading rate, and to calculate a second product of the second loss rate and the second shading rate;

[0174] The total loss rate calculation unit is used to calculate the sum of the first product and the second product to obtain the loss rate.

[0175] Optionally, the weight prediction model is trained by the following modules:

[0176] A sample set acquisition module, used to acquire a sample image set, wherein the sample images in the sample image set are images when there are passengers in the elevator car, and each sample image is marked with a third weight;

[0177] A sample input module, used for inputting the sample image into a weight prediction model to obtain a fourth weight of the passenger;

[0178] a weight loss calculation module, configured to calculate a loss rate using the third weight and the fourth weight;

[0179] The weight prediction model training condition judgment module is used to judge whether the preset training conditions are met; if so, the weight prediction model completion training determination module is executed; if not, the weight prediction model adjustment module is executed;

[0180] A weight prediction model training completion determination module is used to determine that the weight prediction model training is completed;

[0181] The weight prediction model adjustment module is used to adjust the model parameters of the weight prediction model according to the loss rate and return it to the sample input module.

[0182] Optionally, the second score correction module 507 includes:

[0183] a correction coefficient determination unit, configured to calculate a correction coefficient based on the first weight and the second weight;

[0184] The second score calculation unit is used to calculate the product of the correction coefficient and the first score to obtain a second score.

[0185] Optionally, the correction coefficient determination unit includes:

[0186] a weight difference calculation subunit, configured to calculate a difference between the first weight and the second weight;

[0187] A difference determination subunit, used to determine whether the difference is greater than the value 0, if not, execute the second determination module 509, if yes, execute the ratio calculation subunit;

[0188] a ratio calculation subunit, configured to calculate a ratio of the difference to the preset weight when the difference is greater than a preset weight threshold;

[0189] The correction coefficient determination subunit is used to determine the correction coefficient in a preconfigured ratio-coefficient table according to the ratio.

[0190] The device for identifying abnormal objects in an elevator based on a visual neural network provided in an embodiment of the present invention can execute the method for identifying abnormal objects in an elevator based on a visual neural network provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0191] Embodiment 4

[0192] Figure 6 A schematic diagram of the structure of an electronic device 40 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0193] like Figure 6 As shown, the electronic device 40 includes at least one processor 41, and a memory connected to the at least one processor 41, such as a read-only memory (ROM) 42, a random access memory (RAM) 43, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 41 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 42 or the computer program loaded from the storage unit 48 to the random access memory (RAM) 43. In the RAM 43, various programs and data required for the operation of the electronic device 40 can also be stored. The processor 41, the ROM 42, and the RAM 43 are connected to each other through a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0194] A number of components in the electronic device 40 are connected to the I / O interface 45, including: an input unit 46, such as a keyboard, a mouse, etc.; an output unit 47, such as various types of displays, speakers, etc.; a storage unit 48, such as a disk, an optical disk, etc.; and a communication unit 49, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 49 allows the electronic device 40 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0195] The processor 41 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 41 executes the various methods and processes described above, such as a method for identifying abnormal objects in an elevator based on a visual neural network.

[0196] In some embodiments, the method for identifying abnormal objects in an elevator based on a visual neural network can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 48. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 40 via the ROM 42 and / or the communication unit 49. When the computer program is loaded into the RAM 43 and executed by the processor 41, one or more steps of the method for identifying abnormal objects in an elevator based on a visual neural network described above can be performed. Alternatively, in other embodiments, the processor 41 can be configured to execute the method for identifying abnormal objects in an elevator based on a visual neural network in any other appropriate manner (e.g., by means of firmware).

[0197] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0198] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0199] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, 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 foregoing.

[0200] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0201] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0202] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0203] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0204] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for identifying abnormal objects in an elevator based on a visual neural network, characterized in that: Used to identify electric vehicles in elevators, including: When the elevator door is opened, a first camera in the elevator car is controlled to collect a first video and a second camera is controlled to collect a second video, wherein the first camera and the second camera are cameras at different positions in the elevator car; Inputting the first video and the second video into an electric vehicle recognition neural network to obtain a first score of the electric vehicle in the elevator car; Determining whether the first score is greater than a preset first score threshold; If so, determining that there is an electric vehicle in the elevator car and generating an alarm message; If not, when the first score is greater than a preset second score threshold, starting a weighing sensor to obtain a first weight of all objects in the elevator car, and the second score threshold is less than the first score threshold; Inputting at least one of the first video and the second video into a weight prediction model to obtain a second weight of the passenger in the elevator car; Correcting the first score using the first weight and the second weight to obtain a second score; Determining whether the second score is greater than the first score threshold; If so, executing the step of determining that there is an electric vehicle in the elevator car and generating an alarm message; If not, it is determined that there is no electric vehicle in the elevator car.

2. The method according to claim 1, characterized in that The electric vehicle recognition neural network includes a first feature extraction network, a second feature extraction network, a feature fusion network, a first detection head, and a second detection head. The first video and the second video are input into the electric vehicle recognition neural network to obtain a first score of the electric vehicle in the elevator car, including: Inputting the first video into a first feature extraction network, and extracting a first feature map from the first video in the first feature extraction network; Inputting the second video into a second feature extraction network, and extracting a second feature map from the second video in the second feature extraction network; fusing the first feature map and the second feature map in the feature fusion network to obtain a fused feature; The first detection head and the second detection head respectively identify the fusion features to obtain a first initial score of the electric vehicle in the first video and a second initial score of the electric vehicle in the second video; An average of the first initial score and the second initial score is calculated as a first score for the presence of an electric vehicle in the elevator car.

3. The method according to claim 2, characterized in that The electric vehicle recognition neural network is trained in the following way: A first image and a second image of an elevator car taken from different angles are obtained as training images, wherein the first image is annotated with a third fraction of electric vehicles and a first occlusion rate of the electric vehicles, and the second image is annotated with a fourth fraction of electric vehicles and a second occlusion rate of the electric vehicles; Inputting the first image and the second image into an electric vehicle recognition neural network to obtain a fifth score indicating the presence of an electric vehicle in the first image and a sixth score indicating the presence of an electric vehicle in the second image; Calculating the loss rate using the third score, the fourth score, the fifth score, the sixth score, the first occlusion rate, and the second occlusion rate; Determine whether preset training conditions are met; If so, determining that the electric vehicle identification neural network has completed training; If not, adjust the network parameters of the electric vehicle identification neural network according to the loss rate, and return to the step of inputting the first image and the second image into the electric vehicle identification neural network.

4. The method according to claim 3, characterized in that The loss rate is calculated using the third score, the fourth score, the fifth score and the sixth score, including: calculating a first loss rate by using the third score and the fifth score, and calculating a second loss rate by using the fourth score and the sixth score; Calculating a first product of the first loss rate and the first occlusion rate, and calculating a second product of the second loss rate and the second occlusion rate; The loss rate is obtained by calculating the sum of the first product and the second product.

5. The method according to any one of claims 1 to 4, characterized in that: The weight prediction model is trained in the following way: Acquire a sample image set, wherein the sample images in the sample image set are images when there are passengers in the elevator car, and each sample image is marked with a third weight; Inputting the sample image into a weight prediction model to obtain a fourth weight of the passenger; calculating a loss rate using the third weight and the fourth weight; Determine whether preset training conditions are met; If so, determining that the weight prediction model has completed training; If not, the model parameters of the weight prediction model are adjusted according to the loss rate, and the process returns to the step of inputting the sample image into the weight prediction model to obtain the second weight of the passenger.

6. The method according to any one of claims 1 to 4, characterized in that: Correcting the first score by using the first weight and the second weight to obtain a second score includes: calculating a correction factor based on the first weight and the second weight; The product of the correction coefficient and the first fraction is calculated to obtain a second fraction.

7. The method according to claim 6, characterized in that Calculating a correction factor based on the first weight and the second weight includes: calculating a difference between the first weight and the second weight; Determine whether the difference is greater than a value of 0; If not, executing the step of determining that there is no electric vehicle in the elevator car; If so, when the difference is greater than a preset weight threshold, calculating a ratio of the difference to the preset weight; The correction factor is determined from the ratio in a preconfigured ratio-factor table.

8. A device for identifying abnormal objects in an elevator based on a visual neural network, characterized in that: Used to identify electric vehicles in elevators, including: A video acquisition module, used for controlling a first camera in the elevator car to acquire a first video and controlling a second camera to acquire a second video when the elevator door is opened, wherein the first camera and the second camera are cameras at different positions in the elevator car; A first score prediction module, configured to input the first video and the second video into an electric vehicle recognition neural network to obtain a first score of the electric vehicle in the elevator car; A first score determination module, used to determine whether the first score is greater than a preset first score threshold, and if so, execute a first determination module, and if not, execute a first weight acquisition module; A first determination module, used to determine the presence of an electric vehicle in the elevator car and generate an alarm message; a first weight acquisition module, configured to start a weighing sensor to acquire a first weight of all objects in the elevator car when the first score is greater than a preset second score threshold, and the second score threshold is less than the first score threshold; A second weight prediction module, configured to input at least one of the first video and the second video into a weight prediction model to obtain a second weight of the passenger in the elevator car; A second score correction module, configured to correct the first score by using the first weight and the second weight to obtain a second score; A second score determination module, used to determine whether the second score is greater than the first score threshold, and if so, execute the first determination module, and if not, execute the second determination module; The second determining module is used to determine whether there is no electric vehicle in the elevator car.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for identifying abnormal objects in an elevator based on a visual neural network according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for identifying abnormal objects in an elevator based on a visual neural network according to any one of claims 1 to 7 when executed.

Images