A method, device, storage medium and electronic device for sharing bicycle fault feedback

Abstract

The present invention provides a method, device, storage medium and electronic device for fault feedback of shared bicycles. A method for fault feedback of shared bicycles includes: obtaining a video stream of a target parking area; identifying shared bicycles parked in the target parking area based on the video stream; detecting whether a preset fault scenario occurs for the shared bicycles in the target parking area based on the video stream; in the case of a preset fault scenario occurring, recording a fault score of the shared bicycle based on a preset rule; and feedbacking maintenance information based on the fault scores of each shared bicycle in the target parking area, solving the problem of reducing maintenance costs and improving the utilization rate of shared bicycles.

Description

Technical Field

[0001] The present invention relates to the technical field of shared bicycles, and in particular, to a method, device, storage medium and electronic device for feedback of shared bicycle faults. Background Art

[0002] In today's life, shared bicycles are widely distributed in major cities to facilitate people's travel. With their convenience, shared bicycles have brought a lot of convenience to people's lives, and also combined travel and exercise, facilitating people's lives and travel. However, shared bicycles are often damaged, and the management personnel of shared bicycles cannot timely discover which bicycles are damaged and need to be repaired, resulting in a decrease in the frequency of people using shared bicycles. Further, due to the large-scale emergence of shared bicycles, the maintenance cost of shared bicycles has gradually become higher. Especially after many shared bicycles have problems, due to the lack of timely repair reports, it is easy to lead to scenarios of scrapping.

[0003] Therefore, there is an urgent need in the art for a solution to monitor the faults of shared bicycles and timely feedback the fault information of shared bicycles, so as to reduce the maintenance cost and improve the utilization rate of shared bicycles. Summary of the Invention

[0004] In order to reduce the maintenance cost and improve the utilization rate of shared bicycles, embodiments of the present invention provide a method, device, storage medium and electronic device for feedback of shared bicycle faults.

[0005] In a first aspect, an embodiment of the present invention provides a method for feedback of shared bicycle faults, including:

[0006] Obtaining a video stream of a target parking area;

[0007] Identifying the shared bicycles parked in the target parking area based on the video stream;

[0008] Detecting whether a preset fault scenario occurs for the shared bicycles in the target parking area based on the video stream;

[0009] In the case of a preset fault scenario occurring, recording a fault score of the shared bicycle based on a preset rule;

[0010] Feedbacking maintenance information based on the fault scores of the shared bicycles in the target parking area.

[0011] In some embodiments, the identifying the shared bicycles parked in the target parking area based on the video stream includes:

[0012] Detecting whether a single person parks a bicycle in the target parking area in the video stream;

[0013] In response to a single person parking a bicycle in the target parking area, determine whether a preset locking action and / or a preset locking success prompt appears;

[0014] If a preset locking action and / or a preset locking success prompt appears, determine that the bicycle is a shared bicycle.

[0015] In some embodiments, the preset locking action includes an action of sliding the lock of the bicycle to close or an action of using a mobile phone to close the lock; the preset locking success prompt includes a voice prompt of successful locking.

[0016] In some embodiments, when the preset locking action includes an action of sliding the lock of the bicycle to close, determining whether a preset locking action appears includes:

[0017] Identify the lock of the bicycle in the video stream and determine whether the lock is a preset lock;

[0018] In response to the lock being a preset lock, identify the position of the lock and the position of the target joint point corresponding to the preset locking action;

[0019] Detect a first overlap rate between the target joint point and the lock;

[0020] In response to the first overlap rate being greater than a first ratio and maintaining a preset duration, determine that a preset locking action appears.

[0021] In some embodiments, identifying the position of the lock and the position of the target joint point corresponding to the preset locking action includes:

[0022] Locate the position of the lock and the target joint point corresponding to the preset locking action frame by frame;

[0023] Calculate the centroid position of the target joint point, and assign a unique ID to the centroid;

[0024] Calculate the distance between the centroids in two adjacent frames, associate the centroid with the smallest distance in the two adjacent frames, and update the ID of the centroid in the current frame with the ID of the centroid in the previous frame to associate the same centroid in the two adjacent frames;

[0025] The detecting the first overlap rate between the target joint point and the lock includes:

[0026] Detect the first overlap rate between the centroid and the lock.

[0027] In some embodiments, identifying the position of the lock and the position of the target joint point corresponding to the preset locking action further includes:

[0028] If a new target joint point is located, calculate the centroid position of the new target joint point and assign a unique ID to the centroid; and / or

[0029] If the target joint point corresponding to the centroid with an assigned ID cannot be located for a continuous preset number of frames, delete the centroid and its ID.

[0030] In some embodiments, the identifying of the shared bicycles parked in the target parking area based on the video stream includes:

[0031] Detect whether a batch of bicycles of the same color are parked in the target parking area in the video stream;

[0032] In response to a batch of bicycles of the same color being parked in the target parking area, randomly identify the lock of one or more bicycles in the batch;

[0033] If the randomly identified lock is a preset lock, determine that the batch of bicycles are shared bicycles.

[0034] In some embodiments, the preset fault scenarios include:

[0035] The first fault scenario: A person performs a preset unlocking action on a shared bicycle in the target parking area, and after the preset unlocking action ends, the person moves while the shared bicycle does not move;

[0036] The second fault scenario: A person parks a shared bicycle in the target parking area, and within a preset time after parking, performs a preset unlocking action on another shared bicycle and leaves the target parking area;

[0037] The third fault scenario: Within a preset time range, the shared bicycles in the target parking area are not moved within a set duration.

[0038] In some embodiments, in the case of a preset fault scenario occurring, recording the fault score of the shared bicycle based on a preset rule includes:

[0039] For a shared bicycle that has the first fault scenario occur continuously for a preset number of times, record a first fault score once;

[0040] For a shared bicycle that has the second fault scenario occur, record a second fault score once for each occurrence of the second fault scenario;

[0041] For a shared bicycle that has the third fault scenario occur, record a third fault score once for each set duration that it is not moved;

[0042] Wherein, the first fault score > the second fault score > the third fault score.

[0043] In some embodiments, the preset unlocking action in the first failure scenario or the second failure scenario includes an action of unlocking the vehicle using a mobile phone, and the preset unlocking action is detected in the following manner:

[0044] Detect the second overlap rate between the target joint points corresponding to the preset unlocking action and the mobile phone;

[0045] In response to the second overlap rate reaching the second proportion, locate the position of the vehicle lock and the target joint points corresponding to the preset unlocking action frame by frame;

[0046] Calculate the centroid position of the target joint points, and assign a unique ID to the centroid;

[0047] Calculate the distance between the centroids in two adjacent frames, associate the centroids with the smallest distance in two adjacent frames, and update the ID of the centroid in the current frame with the ID of the centroid in the previous frame to associate the same centroid in two adjacent frames;

[0048] Detect the first overlap rate between the centroid and the vehicle lock;

[0049] In response to the first overlap rate being greater than the first proportion and maintaining for a preset duration, determine that the preset unlocking action has occurred.

[0050] In some embodiments, the feedback of maintenance information based on the failure scores of each shared bicycle in the target parking area includes:

[0051] Based on the cumulative failure scores of each shared bicycle in the target parking area, determine the maintenance level of each shared bicycle, where the higher the cumulative failure score, the higher the corresponding maintenance level;

[0052] When the number of shared bicycles in the target parking area whose maintenance level reaches the preset maintenance level reaches the preset number, feedback the maintenance information; where the maintenance information includes the information of shared bicycles with different maintenance levels in the target parking area.

[0053] In some embodiments, the feedback of maintenance information based on the failure scores of each shared bicycle in the target parking area includes:

[0054] Based on the cumulative failure scores of each shared bicycle in the target parking area, determine the maintenance level of each shared bicycle, where the higher the cumulative failure score, the higher the corresponding maintenance level;

[0055] Calculate the sum of the cumulative failure scores of each shared bicycle in the target parking area,

[0056] When the sum of the cumulative failure scores reaches a preset score, maintenance information is fed back; wherein, the maintenance information includes the information of shared bicycles with different maintenance levels in the target parking area.

[0057] In some embodiments, the method further includes:

[0058] Collecting the data of shared bicycles in the target parking area and forming a visual report according to a preset period;

[0059] Wherein, the data of shared bicycles includes at least one of the number of shared bicycles parked in the target parking area within a preset period, the parking time of the shared bicycles, and the gender of the persons parking the shared bicycles.

[0060] In some embodiments, the method further includes:

[0061] According to the collected data of shared bicycles, providing a suggestion on the optimal number of shared bicycles parked in the target parking area; and / or

[0062] Collecting the number of shared bicycles parked outside the target parking area, determining whether the demarcation of the target parking area is reasonable, and providing a suggestion on the demarcation of the target parking area when the demarcation of the target parking area is unreasonable.

[0063] In some embodiments, the providing a suggestion on the optimal number of shared bicycles parked in the target parking area according to the collected data of shared bicycles includes:

[0064] Counting the traffic flow in the target parking area;

[0065] Based on the average value of the traffic flow, determining the optimal number of shared bicycles parked in the target parking area;

[0066] Wherein, the optimal number of shared bicycles parked includes the optimal number of shared bicycles parked on weekdays and / or the optimal number of shared bicycles parked on rest days.

[0067] In some embodiments, the collecting the data of shared bicycles in the target parking area and forming a visual report according to a preset period includes:

[0068] Collecting the data of shared bicycles in the target parking area within a preset period;

[0069] Converting the data of shared bicycles into a target data format;

[0070] Performing missing value processing on the data of shared bicycles in the target data format to complete the missing fields;

[0071] Perform outlier processing on the shared bicycle data after missing value processing, and the outlier processing includes deleting duplicate data;

[0072] Generate a visualization report based on the shared bicycle data after outlier processing.

[0073] In a second aspect, an embodiment of the present invention provides a shared bicycle fault feedback device, including:

[0074] An acquisition module for acquiring a video stream of a target parking area;

[0075] An identification module for identifying shared bicycles parked in the target parking area based on the video stream;

[0076] A detection module for detecting whether a preset fault scenario occurs for the shared bicycles in the target parking area based on the video stream;

[0077] A recording module for recording the fault scores of the shared bicycles based on a preset rule when a preset fault scenario occurs;

[0078] A feedback module for feedbacking maintenance information based on the fault scores of the shared bicycles in the target parking area.

[0079] In a third aspect, an embodiment of the present invention provides a storage medium, on which a computer program is stored, and when the computer program is executed by one or more processors, the shared bicycle fault feedback method as described in the first aspect is implemented.

[0080] In a fourth aspect, an embodiment of the present invention provides an electronic device, including a memory and one or more processors, and a computer program is stored on the memory, and when the computer program is executed by the one or more processors, the shared bicycle fault feedback method as described in the first aspect is implemented.

[0081] Compared with the prior art, one or more embodiments of the present invention can at least bring the following beneficial effects:

[0082] The present invention obtains the video stream of the target parking area in real time, identifies the shared bicycles parked in the target parking area based on the video stream, and when there are shared bicycles parked in the target parking area, detects whether a preset fault scenario occurs for the shared bicycles in the target parking area. When a preset fault scenario occurs, the fault score of the shared bicycle is recorded based on a preset rule. Furthermore, based on the fault scores of the shared bicycles in the target parking area, maintenance information is fed back in a timely manner, realizing the fault monitoring of the shared bicycles and being able to feed back the maintenance information of the shared bicycles in a timely manner according to the fault scores, enabling the managers of the shared bicycles to obtain the fault conditions of the shared bicycles in a timely manner and arranging maintenance according to the feedback conditions in a timely manner to reduce the later maintenance cost and improve the utilization rate of the shared bicycles. At the same time, by regularly analyzing the data of the bicycle information, a visual analysis report can be formed, as well as reasonable suggestions for the number of bicycles that should be stored in this area and whether the parking area is user-friendly, achieving the effect of continuously optimizing the parking area and improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0084] Figure 1 is a schematic flow chart of a method for feedback of shared bicycle faults provided by an embodiment of the present invention;

[0085] Figure 2 is a schematic flow chart of a method for identifying shared bicycles parked in a target parking area based on a video stream provided by an embodiment of the present invention;

[0086] Figure 3 is another schematic flow chart of a method for identifying shared bicycles parked in a target parking area based on a video stream provided by an embodiment of the present invention;

[0087] Figure 4 is a schematic flow chart of a method for determining whether a preset unlocking action occurs provided by an embodiment of the present invention;

[0088] Figure 5 is a schematic flow chart of a method for identifying the position of the vehicle lock and the position of the target joint point corresponding to the preset unlocking action provided by an embodiment of the present invention;

[0089] Figure 6 is another schematic flow chart of a method for feedback of shared bicycle faults provided by an embodiment of the present invention;

[0090] Figure 7It is a schematic diagram of a shared bicycle fault feedback device provided by an embodiment of the present invention. Detailed implementation manners

[0091] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Usually, the components of the embodiments of the present invention described and illustrated herein can be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of the present invention provided in the accompanying drawings below is not intended to limit the scope of the present invention to be protected, but only represents the selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0092] Due to the large-scale emergence of shared bicycles, the maintenance cost of shared bicycles has gradually become higher. Especially after many shared bicycles have malfunctioned, due to the lack of timely repair reports, scenarios of scrapping have occurred. Therefore, there is an urgent need in the art for a solution to monitor the faults of shared bicycles and timely feedback the fault information of shared bicycles, so as to reduce the maintenance cost and improve the utilization rate of shared bicycles. In order to reduce the maintenance cost and improve the utilization rate of shared bicycles, an embodiment of the present invention provides a shared bicycle fault feedback method, device, storage medium and electronic device.

[0093] Embodiment 1

[0094] Figure 1 shows a schematic flowchart of a shared bicycle fault feedback method. As Figure 1 shown, this embodiment provides a shared bicycle fault feedback method, including steps S110 to S150:

[0095] Step S110, obtain the video stream of the target parking area.

[0096] Step S120, identify the shared bicycles parked in the target parking area based on the video stream.

[0097] Step S130, detect whether the shared bicycles in the target parking area have a preset fault scenario based on the video stream.

[0098] Step S140, in the case of a preset fault scenario, record the fault score of the shared bicycle based on a preset rule.

[0099] Step S150, feedback the repair information based on the fault scores of the shared bicycles in the target parking area.

[0100] In this embodiment, the video stream of the target parking area is obtained in real time, the shared bicycles parked in the target parking area are identified based on the video stream, and when there are shared bicycles parked in the target parking area, it is detected whether a preset fault scenario occurs for the shared bicycles in the target parking area. When a preset fault scenario occurs, the fault score of the shared bicycle is recorded based on a preset rule, and then, based on the fault scores of the shared bicycles in the target parking area, maintenance information is fed back in a timely manner, realizing the fault monitoring of the shared bicycles and being able to feed back the maintenance information of the shared bicycles in a timely manner according to the fault scores, enabling the managers of the shared bicycles to obtain the fault conditions of the shared bicycles in a timely manner and arranging maintenance in a timely manner according to the feedback situation, so as to reduce the later maintenance cost and improve the utilization rate of the shared bicycles.

[0101] In practical applications, the shared bicycles parked in the target parking area can be identified every 20 frames, and each identification is compared with the previous 20 frames to identify newly added shared bicycles. When a shared bicycle parked in the target parking area is identified based on the video stream, a unique ID can be assigned to the identified shared bicycle and the parking time can be recorded. The shared bicycles may be parked individually or in batches. In this embodiment, implementation methods for identifying the shared bicycles parked in the target parking area based on the video stream are provided in two cases respectively.

[0102] As Figure 2 shown, in one implementation manner, step S120 of identifying the shared bicycles parked in the target parking area based on the video stream may include steps S210 to S240:

[0103] Step S210, detect whether there are bicycles of the same color parked in batches in the target parking area in the video stream; in response to the appearance of bicycles of the same color parked in batches in the target parking area, execute step S220; in response to the non-appearance of bicycles of the same color parked in batches in the target parking area, the step S110 of obtaining the video stream of the target parking area may be executed, and continue to detect whether there are bicycles of the same color parked in batches in the target parking area in the video stream to execute Figure 2 the process shown. It should be understood that in the application of this method, while detecting whether there are bicycles of the same color parked in batches in the target parking area in the video stream, it is possible to synchronously detect whether there are single bicycles parked in the target parking area in the video stream to execute Figure 3 the process shown.

[0104] Step S220, randomly identify the locks of one or more bicycles in this batch of bicycles;

[0105] Step S230, determine whether the randomly identified lock is a preset lock; if the randomly identified lock is a preset lock, execute step S240;

[0106] Step S240: Determine that this batch of bicycles are shared bicycles. If the randomly identified bicycle lock is not the preset bicycle lock, then this batch of bicycles are not shared bicycles.

[0107] In actual situations, the bicycles placed in batches are usually shared bicycles, and the batch can be, but is not limited to, more than 10 bicycles. Therefore, in this implementation method, by randomly identifying whether the lock of one or more bicycles in the batch of bicycles placed in batches is the preset bicycle lock, it is possible to determine whether this batch of bicycles are shared bicycles at one time, without having to identify them one by one.

[0108] As Figure 3 shown, in another implementation method, step S120 for identifying shared bicycles parked in the target parking area based on the video stream can further include steps S310 to S330:

[0109] Step S310: Detect whether a single person parks a bicycle in the target parking area in the video stream; in response to a single person parking a bicycle in the target parking area, execute step S320; in response to no single person parking a bicycle in the target parking area, the step S110 of obtaining the video stream of the target parking area can be executed to continue detecting whether a single person parks a bicycle in the target parking area to execute Figure 3 the process shown. It should be understood that in the application of this method, while detecting whether a single person parks a bicycle in the target parking area in the video stream, it is possible to synchronously detect whether a batch of bicycles of the same color are parked in the target parking area in the video stream to execute Figure 2 the process shown.

[0110] Step S320: Determine whether a preset lock - closing action and / or a preset lock - closing success prompt appear; if a preset lock - closing action and / or a preset lock - closing success prompt appear, execute step S330;

[0111] Step S330: Determine that this bicycle is a shared bicycle. It should be understood that if a preset lock - closing action and / or a preset lock - closing success prompt do not appear, then this bicycle is not a shared bicycle.

[0112] Since in actual applications, some shared bicycles can be locked by sliding the bicycle lock, some can be locked through the mobile phone APP, and some may be locked through other lock - closing actions, the specific action of the preset lock - closing action is not limited in this embodiment. The preset lock - closing action can include, but is not limited to, the action of sliding and closing the bicycle lock or the action of closing the bicycle lock using the mobile phone. The preset lock - closing success prompt can include, but is not limited to, the voice prompt of lock - closing success. Determine whether a preset lock - closing success prompt appears through voice recognition. For example, voice prompts such as "Thank you for using XX bicycle" or "Lock - closing successful". Of course, the voice prompts for lock - closing success of different shared bicycles may be different, and the specific voice prompt is not limited in this embodiment.

[0113] In some cases, the preset locking action can be detected and recognized by the characteristics of the action. Preferably, a framework of an object detection algorithm can be built using a recurrent neural network to determine whether the preset locking action occurs. Specifically, the co-occurrence characteristics of the joint points in the action can be introduced into the design of the recurrent neural network and used as a constraint for network parameter learning to optimize the recognition performance. The set of specific joint points of the skeleton corresponding to the vehicle lock and the preset vehicle lock action can use the OpenPose algorithm to detect the target joint points in the image, including the wrists. By detecting the degree of coincidence between the target joint points and the position of the vehicle lock, it is determined whether the preset locking action occurs. The closer the position coordinates of the target joint points and the vehicle lock are, the more the wrist area overlaps with the rectangular area of the vehicle lock. The action of closing the vehicle lock is generally a continuous process. Therefore, after the action is detected in the first frame, when the action duration exceeds a certain length (for example, 2 s), it indicates that the person is performing the action of closing the vehicle lock.

[0114] In one implementation, in the case where the preset locking action includes the action of sliding and closing the vehicle lock, determining whether the preset locking action occurs in step S320, as Figure 4 shown, may include:

[0115] Step S410, identifying the vehicle lock of the bicycle in the video stream;

[0116] Step S420, determining whether the vehicle lock is a preset vehicle lock; in response to the vehicle lock being a preset vehicle lock, execute step S430; in response to the vehicle lock not being a preset vehicle lock, then the bicycle is not a shared bicycle;

[0117] Step S430, identifying the position of the vehicle lock and the position of the target joint points corresponding to the preset locking action;

[0118] Step S440, detecting the first overlap rate between the target joint points and the vehicle lock;

[0119] Step S450, determining whether the first overlap rate is greater than the first proportion and maintaining the preset duration; in response to the first overlap rate being greater than the first proportion and maintaining the preset duration, execute step S460; in response to the first overlap rate not being greater than the first proportion, or the first overlap rate being greater than the first proportion but not maintaining the preset duration, determine that the preset locking action does not occur.

[0120] Step S460, determining that the preset locking action occurs.

[0121] In practical applications, the first proportion can be, but is not limited to, 50%, and the preset duration can be, but is not limited to, 2 s. By identifying the lock of a shared bicycle whose lock is not preset, and the first overlap rate between the target joint point and the lock, the detection and recognition of the preset lock closing action are realized. When the first overlap rate is greater than the first proportion, it is the instant when the preset lock closing action is detected. When this action lasts for the preset duration, it indicates that the preset lock closing action has indeed occurred. The above first overlap rate includes the percentage of the area of the overlapping region between the target joint point and the lock in the area of the lock, or the percentage of the area of the overlapping region between the target joint point and the lock in the area of the target joint point.

[0122] It should be understood that through adaptive adjustment, steps S410 to S460 can also be equally applicable to the case where the preset lock closing action includes the action of closing the lock using a mobile phone. It can be detected by identifying the mobile phone and the target joint point (wrist) in the video stream and detecting the second overlap rate between the wrist and the mobile phone. The principle is the same as before and will not be elaborated here.

[0123] As Figure 5 shown, in some implementation manners, step S430 of identifying the position of the lock and the position of the target joint point corresponding to the preset lock closing action may further include:

[0124] Step S430a: Locate the position of the lock and the target joint point corresponding to the preset lock closing action frame by frame;

[0125] Step S430b: Calculate the centroid position of the target joint point and assign a unique ID to the centroid;

[0126] Step S430c: Calculate the distance between the centroids in two adjacent frames, associate the centroids with the smallest distance in two adjacent frames, and update the ID of the centroid in the current frame with the ID of the centroid in the previous frame to associate the same centroid in two adjacent frames.

[0127] Correspondingly, step S440 of detecting the first overlap rate between the target joint point and the lock may include:

[0128] Step S440a: Detect the first overlap rate between the centroid and the lock.

[0129] Furthermore, step S430 of identifying the position of the lock and the position of the target joint point corresponding to the preset lock closing action may further include:

[0130] If a new target joint point is located, calculate the centroid position of the new target joint point and assign a unique ID to the centroid; and / or

[0131] If the target joint point corresponding to the centroid with an assigned ID cannot be located for a continuous set number of frames, delete the centroid and its ID.

[0132] In practical applications, the set number of frames can be, but is not limited to, 20 frames.

[0133] In some cases, the target joint points corresponding to the above-mentioned preset locking actions include the wrist. Locate the position of the bike lock and the position of the person's wrist in each frame of the video stream, calculate the centroid of the wrist, assign a unique ID to each centroid, calculate the new centroid in the next frame, calculate the distance between the centroids of the current frame and the previous frame, and associate the centroids in the two frames based on the principle of the minimum distance. If there is a correlation between the centroids in the two frames, update the new centroid with the ID of the centroid in the previous frame to ensure the uniqueness of the centroid ID. Detect the first overlap rate between the centroid and the bike lock. If the first overlap rate is greater than the first proportion and maintains for a preset duration, it is determined that the preset locking action has occurred. For example, if the first overlap rate is greater than 50% and lasts for 2 s, it indicates that the preset locking action has occurred. It should be understood that calculating the centroid position can simplify the calculation of the first overlap rate, and the centroid is not limited to being a single point. It can be a region representing the target joint point for calculating the first overlap rate.

[0134] Step S460, determine that the preset locking action has occurred.

[0135] If no correlation is found between the centroids in the two frames, that is, there is no pair of centroids with the minimum distance in two adjacent frames, the target joint points identified in the current frame are those of a new person, and the corresponding calculated centroid is a newly added centroid. Assign a unique ID to this new centroid and associate the same centroid in each frame according to the above process. If it is detected that a certain target joint point leaves the video stream for the set number of frames, for example, the target joint point cannot be recognized for 20 consecutive frames, it is considered that the corresponding person has left the target parking area, and then remove the centroid corresponding to the target joint point and its ID.

[0136] On the basis of identifying the shared bikes parked in the target parking area based on the video stream, further detect whether the shared bikes in the target parking area have preset fault scenarios based on the video stream. In practical applications, only a small number of shared bikes have preset fault scenarios, or the occurrence of preset fault scenarios occurs occasionally in the case of free rides or within the riding period. If immediate repair is reported, it will increase the maintenance cost or result in false fault reports. Therefore, in the case of a preset fault scenario, record the fault score of the shared bike based on the preset rules, and feedback the maintenance information based on the fault scores of each shared bike in the target parking area, so as to avoid the increase in maintenance costs and also avoid the situation of false repair reports due to preset fault scenarios that occur occasionally due to free rides or within the riding period.

[0137] In some embodiments, the above-mentioned preset fault scenarios may include:

[0138] The first failure scenario: A person performs a preset unlocking action on a shared bicycle in the target parking area, and after the preset unlocking action ends, the person moves while the shared bicycle does not move;

[0139] The second failure scenario: A person parks a shared bicycle in the target parking area, and within a preset time after parking, performs a preset unlocking action on another shared bicycle and leaves the target parking area;

[0140] The third failure scenario: Within a preset time range, the shared bicycle in the target parking area has not been moved within a set duration.

[0141] It should be understood that the above three failure scenarios are distinguished according to the possible degree of failure.

[0142] In some cases, the preset unlocking action in the first failure scenario or the second failure scenario includes the action of opening the vehicle lock using a mobile phone. The scanning code action and some specific joint points of the skeleton form a set. The specific joint points include the wrist. The smaller the distance between the detected joint point and the position of the vehicle lock, the closer the position coordinates of the wrist area and the rectangular area of the vehicle lock. The scanning code action is a continuous process. Therefore, when it exceeds 2s after detecting the action from the first frame, it indicates that the person is performing the scanning code action.

[0143] Therefore, in one implementation, the preset unlocking action can be detected in the following manner:

[0144] Step a: Detect the second overlap rate between the target joint point corresponding to the preset unlocking action and the mobile phone; in response to the second overlap rate reaching the second proportion, execute step b;

[0145] Step b: Locate the position of the vehicle lock and the target joint point corresponding to the preset unlocking action frame by frame;

[0146] Step c: Calculate the centroid position of the target joint point and assign a unique ID to the centroid;

[0147] Step d: Calculate the distance between the centroids in two adjacent frames, associate the centroids with the smallest distance in two adjacent frames, and update the ID of the centroid in the current frame using the ID of the centroid in the previous frame to associate the same centroid in two adjacent frames;

[0148] Step e: Detect the first overlap rate between the centroid and the vehicle lock;

[0149] Step f: Determine whether the first overlap rate is greater than the first proportion and maintains a preset duration; in response to the first overlap rate being greater than the first proportion and maintaining a preset duration, determine that the preset unlocking action occurs. In response to the first overlap rate not being greater than the first proportion or the first overlap rate being greater than the first proportion but not maintaining the preset duration, determine that the preset unlocking action does not occur.

[0150] In practical applications, the second proportion can be, but is not limited to, 90%, and the preset duration can be, but is not limited to, 2 s. The above-mentioned second overlap rate includes the percentage of the area of the overlapping region between the target joint point and the mobile phone in the area of the mobile phone, or the percentage of the area of the overlapping region between the target joint point and the mobile phone in the area of the target joint point. When the second overlap rate reaches more than 90%, it indicates that a person is holding the mobile phone. When the first overlap rate between the centroid and the bike lock reaches more than 50% and lasts for 2 s, it indicates that the person is scanning the code to unlock the bike. If no correlation is found for the centroids in two frames, that is, there is no pair of centroids with the smallest distance in two adjacent frames, then the target joint point identified in the current frame belongs to a new person, and the corresponding centroid calculated is the newly added centroid. A unique ID is assigned to this new centroid, and the centroid is associated with the same centroid in each frame according to the above process. If it is detected that a certain target joint point has left the video stream for a set number of frames, for example, the target joint point cannot be recognized for 20 consecutive frames, then it is considered that the corresponding person has left the target parking area, and the centroid and its ID corresponding to the target joint point are removed.

[0151] In an example of the first failure scenario, the preset unlocking action includes a code-scanning action. The closer the position coordinates of the code-scanning action are to the position coordinates of the bike lock of the shared bike, the higher the overlap rate between the area of the code-scanning action and the rectangular area of the shared bike, or the closer the distance between the rectangular area of the code-scanning action and the rectangular area of the shared bike. Therefore, the target shared bike targeted by the code-scanning action can be determined. After the code-scanning action ends; when the person moves, the targeted shared bike does not move. This scenario indicates that the shared bike has an abnormal situation of unlocking failure, and the degree of the failure is relatively high. Of course, in some cases, this scenario may also occasionally occur due to specific situations such as problems with the mobile phone software or Bluetooth connection when the shared bike is not faulty.

[0152] In an example of the second failure scenario, the preset time is 5 s. After the person parks the bike in the target parking area, the timing starts. If the person scans other shared bikes and leaves the corresponding video stream within 5 minutes, the bike can be used in this scenario, but it may affect the user experience, and the degree of the failure is average. Of course, in some cases, this scenario may also be a situation of changing bikes due to the expiration of the free riding time limit when the shared bike is not faulty.

[0153] In an example of the third failure scenario, within the preset time range set by the user for the high-frequency parking time range of the target parking area or based on historical data analysis for the time range when the passenger flow exceeds a certain value (for example, 1000 people / hour), if the shared bike has not been moved for every hour within such a preset time range, it proves that the shared bike may have other damages that affect the user experience other than the aforementioned two failure scenarios, such as air leakage or deformation of the wheels, poor brakes, incomplete pedals, damaged seat cortex, or other problems.

[0154] Therefore, in some implementations, in the case of a preset fault scenario, record the fault score of the shared bicycle based on a preset rule, so as to avoid reporting a repair for any one of the above fault scenarios occurring once, wasting repair costs and resources, and reducing the repair efficiency. Therefore, recording the fault score of the shared bicycle based on a preset rule may include:

[0155] (1) For the shared bicycles that have the first fault scenario occur continuously for a preset number of times, record a first fault score once; for example, the preset number of times is 10 times, the first fault score is 5 points, and the first fault scenario is the scenario corresponding to 5 points. When a certain shared bicycle has the first fault scenario occur continuously 10 times, record 5 points for this shared bicycle once. Scoring is only performed after the continuous preset number of times, avoiding misjudging an occasionally occurring scenario as needing repair.

[0156] (2) For the shared bicycles that have the second fault scenario occur, record a second fault score each time the second fault scenario occurs; for example, the second fault score is 3 points, and the second fault scenario is the scenario corresponding to 3 points. After a certain shared bicycle is parked in the target parking area each time, if any other shared bicycle is scanned and left the video stream corresponding picture within 5 minutes, record 3 points for the parked shared bicycle once.

[0157] (3) For the shared bicycles that have the third fault scenario occur, record a third fault score every time a set duration has passed without being moved; for example, the set duration is 1 hour, and the preset time range is 10:00 - 18:00. Within this preset time range, for each additional hour that the shared bicycle is not moved, count 1 point.

[0158] Among them, the first fault score > the second fault score > the third fault score.

[0159] It should be understood that the above-mentioned various fault scores can be accumulated. Each shared bicycle may have different fault scenarios. If the preset rule is satisfied, record the corresponding score.

[0160] In practical applications, after recording the fault scores of each shared bicycle in the target parking area, the cumulative fault score of each shared bicycle can be counted, and its repair level can be evaluated, and then corresponding repair information can be fed back under certain conditions. Therefore, in some implementations, step S150 feeds back repair information based on the fault scores of each shared bicycle in the target parking area, and may include:

[0161] Step S150 - a1: Determine the repair level of each shared bicycle based on the cumulative fault score of each shared bicycle in the target parking area. Among them, the higher the cumulative fault score, the higher the corresponding repair level;

[0162] Step S150-a2: When the number of shared bikes with a maintenance level reaching the preset maintenance level in the target parking area reaches the preset quantity, feedback maintenance information; wherein, the maintenance information includes the information of shared bikes with different maintenance levels in the target parking area.

[0163] In one example, the first fault score is 5 points, the second fault score is 3 points, the third fault score is 1 point, the preset maintenance level is severe, the preset quantity is 10 bikes, and the corresponding relationship between the cumulative fault score and the maintenance level is as follows:

[0164] Below 10 points, mild;

[0165] 10 - 20 points, moderate;

[0166] Above 20 points, severe.

[0167] Based on such a corresponding relationship, the maintenance level of each shared bike in the target parking area can be determined according to the cumulative fault score. Furthermore, when the number of shared bikes with a severe maintenance level reaches 10, feedback the maintenance information. The feedback maintenance information includes not only the information of shared bikes with a severe maintenance level but also the shared bikes with a mild or moderate maintenance level in the target parking area. After synchronously sending this maintenance information to the shared bike manager, the shared bike manager makes maintenance arrangements according to the feedback maintenance information.

[0168] In practical applications, after recording the fault scores of each shared bike in the target parking area, the cumulative fault scores of each shared bike can also be summarized, and then corresponding maintenance information is feedback under certain conditions. Therefore, in some other implementation manners, step S150 of feedbacking maintenance information based on the fault scores of each shared bike in the target parking area may include:

[0169] Step S150-b1: Determine the maintenance level of each shared bike based on the cumulative fault scores of each shared bike in the target parking area, wherein, the higher the cumulative fault score, the higher the corresponding maintenance level;

[0170] Step S150-b2: Calculate the sum of the cumulative fault scores of each shared bike in the target parking area.

[0171] Step S150-b3: When the sum of the cumulative fault scores reaches the preset score, feedback the maintenance information; wherein, the maintenance information includes the information of shared bikes with different maintenance levels in the target parking area.

[0172] In one example, the first fault score is 5 points, the second fault score is 3 points, the third fault score is 1 point, the preset maintenance level is severe, the preset quantity is 10 bikes, the preset score is 200 points, and the corresponding relationship between the cumulative fault score and the maintenance level is as follows:

[0173] Below 10 points, mild;

[0174] 10 - 20 points, moderate;

[0175] Above 20 points, severe.

[0176] Based on such a correspondence, sum up the cumulative failure scores of each shared bicycle in the target parking area. When the sum of the cumulative failure scores reaches 200 points, feedback maintenance information. The feedback maintenance information includes shared bicycles with different maintenance levels of mild, moderate, and severe in the target parking area. After synchronously sending this maintenance information to the shared bicycle manager, the shared bicycle manager makes maintenance arrangements according to the feedback maintenance information.

[0177] In some embodiments, as Figure 6 shown, this method further includes:

[0178] Step S160: Collect shared bicycle data in the target parking area and form a visualization report according to a preset period; wherein, the shared bicycle data includes at least one of the number of shared bicycles parked in the target parking area within the preset period, the parking time of the shared bicycles, and the gender of the person parking the shared bicycles.

[0179] Furthermore, collecting shared bicycle data in the target parking area and forming a visualization report according to a preset period may include steps S160a to S160e:

[0180] Step S160a: Collect shared bicycle data in the target parking area within the preset period;

[0181] In practical applications, the preset period can be, but is not limited to, one week, one month, or one quarter, etc., and can be set according to actual needs. Collecting shared bicycle data in the target parking area within the preset period can be stored in tabular form. The fields in the table can include the ID of the target parking area and at least one of the number of shared bicycles parked in the target parking area, the parking time of the shared bicycles, and the gender of the person parking the shared bicycles. The parking time of the shared bicycles is the time point when the shared bicycles are parked in the target parking area.

[0182] Step S160b: Convert the shared bicycle data into the target data format;

[0183] Specifically, before forming the visualization report, view all the collected shared bicycle data. When the format of the collected shared bicycle data is not the target data format, the collected shared bicycle data should be converted into the required data type. For example, the format of the collected parking time of the shared bicycles is int, and the target data format is the time format.

[0184] Step S160c: Process the missing values in the shared bike data in the target data format to complete the missing fields;

[0185] In some cases, the collected shared bike data has missing fields. To ensure the effectiveness of data analysis, field completion is required. For example, if the user gender is a null value, a special value (such as -1) can be assigned, thus not affecting data analysis.

[0186] Step S160d: Process the outliers in the shared bike data after missing value processing. The outlier processing includes deleting duplicate data;

[0187] In some cases, for the same shared bike ID in the collected shared bike data, multiple copies of data are repeatedly recorded, then there are outliers. At this time, only the first recorded data can be retained, and the duplicate data can be deleted.

[0188] Step S160e: Generate a visualization report based on the shared bike data after outlier processing.

[0189] Through the above Steps S160a to S160e, the collected dirty data is converted into cleaned data that meets the data quality. By analyzing the data, a visualization report within a preset period can be formed.

[0190] In this embodiment, by regularly analyzing the shared bike data, it is possible to understand the number of shared bikes parked in the target parking area, the parking time of the shared bikes, and the gender of the people parking the shared bikes within a preset period, and based on these data, a visualization report is formed to facilitate the management of the shared bikes in the target parking area by the manager. The forms of the visualization report include but are not limited to visualization charts such as line charts, bar charts, column charts, tables, and scatter plots.

[0191] Furthermore, this method may further include:

[0192] Step S170: Provide a suggestion on the optimal number of shared bikes to be parked in the target parking area according to the collected shared bike data; and / or

[0193] Step S180: Collect the number of shared bikes parked outside the target parking area, determine whether the demarcation of the target parking area is reasonable, and provide a suggestion on the demarcation of the target parking area in the case where the demarcation of the target parking area is unreasonable.

[0194] In this embodiment, based on the collected shared bicycle data, providing a suggestion on the optimal number of shared bicycles to be parked in the target parking area enables the administrator to arrange the number of parked bicycles according to the suggestion. In some cases, the design of the parking area for shared bicycles may not be user-friendly enough, which reduces the frequency of people using shared bicycles and the user experience. Therefore, whether the demarcation of the target parking area is reasonable and providing a demarcation suggestion for the target parking area in the case where the demarcation of the target parking area is unreasonable can effectively solve this problem. By continuously adjusting the demarcation of the parking area based on the suggestion, the user experience can be continuously improved and the utilization rate of shared bicycles can be increased.

[0195] In some implementation manners, based on the collected shared bicycle data, providing a suggestion on the optimal number of shared bicycles to be parked in the target parking area may include:

[0196] Step S170a: Count the traffic flow in the target parking area;

[0197] Step S170b: Determine the optimal number of shared bicycles to be parked in the target parking area based on the average value of the traffic flow; wherein, the optimal number of shared bicycles to be parked includes the optimal number of shared bicycles on weekdays and / or the optimal number of shared bicycles on rest days.

[0198] Specifically, taking a preset period of one month as an example, count the daily traffic flow in the target parking area within one month, calculate the average traffic flow on rest days and / or the average traffic flow on weekdays within this month, and based on the average traffic flow on rest days and / or the average traffic flow on weekdays, determine the optimal number of shared bicycles to be parked on weekdays and / or the optimal number of shared bicycles to be parked on rest days.

[0199] In one example, collecting the number of shared bicycles parked outside the target parking area, determining whether the demarcation of the target parking area is reasonable and providing a demarcation suggestion for the target parking area in the case where the demarcation of the target parking area is unreasonable may include:

[0200] Collect the number of shared bicycles parked outside the target parking area. If the average daily value within a preset period (one month) exceeds a preset value (e.g., 20), it is determined that the demarcation of the target parking area is unreasonable and cannot meet the parking demand; in the case where the demarcation of the target parking area is unreasonable, provide a demarcation suggestion for the target parking area, including determining the area that needs to be increased in the current target parking area according to the area required for parking each shared bicycle and the average value of the number of shared bicycles parked outside the target parking area within the preset period, and further determining the setting orientation of the area that needs to be increased according to the historical parking orientation preference. It should be understood that the area that needs to be increased can be expanded on the basis of the original target parking area, or a new independent parking area can be set separately to relieve the original target parking area. Among them, the historical parking orientation preference can be, but is not limited to, the orientation where the most shared bicycles are parked outside the target parking area detected based on the video stream within the preset period.

[0201] Embodiment 2

[0202] Figure 7 Shows a schematic diagram of a shared bicycle fault feedback device, as Figure 7 shown, this embodiment provides a shared bicycle fault feedback device, including the following modules:

[0203] An acquisition module 710, configured to acquire the video stream of the target parking area;

[0204] An identification module 720, configured to identify the shared bicycles parked in the target parking area based on the video stream;

[0205] A detection module 730, configured to detect whether a preset fault scenario occurs for the shared bicycles in the target parking area based on the video stream;

[0206] A recording module 740, configured to record the fault score of the shared bicycle based on a preset rule in the case of a preset fault scenario;

[0207] A feedback module 750, configured to feedback maintenance information based on the fault scores of each shared bicycle in the target parking area.

[0208] Those skilled in the art should understand that the above-mentioned modules or steps can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. Optionally, they can be implemented by program codes executable by the computing device, so that they can be stored in the storage device and executed by the computing device, or they can be separately made into individual integrated circuit modules, or multiple modules or steps among them can be made into a single integrated circuit module to implement. The present invention is not limited to any defined combination of hardware and software.

[0209] Embodiment 3

[0210] This embodiment provides a storage medium, on which a computer program is stored. When the computer program is executed by one or more processors, the method for feedback on shared bicycle failures in Embodiment 1 is implemented.

[0211] In this embodiment, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM for short), electrically erasable programmable read-only memory (EEPROM for short), erasable programmable read-only memory (EPROM for short), programmable read-only memory (PROM for short), read-only memory (ROM for short), magnetic memory, flash memory, a magnetic disk, or an optical disc. For the content of the method, please refer to Embodiment 1 and will not be elaborated here.

[0212] Embodiment 4

[0213] This embodiment provides an electronic device, including a memory and one or more processors. A computer program is stored on the memory. When the computer program is executed by one or more processors, the method for feedback on shared bicycle failures in Embodiment 1 is implemented.

[0214] In this embodiment, the processor can be implemented by an application specific integrated circuit (ASIC for short), a digital signal processor (DSP for short), a digital signal processing device (DSPD for short), a programmable logic device (PLD for short), a field programmable gate array (FPGA for short), a controller, a microcontroller, a microprocessor, or other electronic components, and is used to execute the method in the above embodiment. The method implemented when the computer program running on the processor is executed can refer to the specific embodiments of the method provided in the foregoing embodiments of the present invention and will not be elaborated here.

[0215] In several embodiments provided by the embodiments of the present invention, it should be understood that the disclosed systems and methods can also be implemented in other ways. The system and method embodiments described above are merely illustrative.

[0216] It should be noted that in this document, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, article or device including the said element.

[0217] Although the disclosed embodiments of the present invention are as above, the above content is only an embodiment adopted for the convenience of understanding the present invention, and is not intended to limit the present invention. Any person skilled in the art within the technical field to which the present invention pertains may make any modifications and changes in the form of implementation and details without departing from the spirit and scope disclosed by the present invention. However, the scope of patent protection of the present invention shall still be subject to the scope defined by the appended claims.

Claims

1. A method for feedback on the failure of a shared bicycle, characterized in that, Including: Obtain the video stream of the target parking area; Identify shared bicycles parked in the target parking area based on the video stream; Detect whether a preset fault scenario occurs for the shared bicycles in the target parking area based on the video stream; In the case of a preset fault scenario, record the fault score of the shared bicycle based on a preset rule; Feedback maintenance information based on the fault scores of the shared bicycles in the target parking area; The identifying the shared bicycles parked in the target parking area based on the video stream includes: Detect whether a single person parks a bicycle in the target parking area in the video stream; In response to a single person parking a bicycle in the target parking area, determine whether a preset lock-closing action and / or a preset lock-closing success prompt appears; If a preset lock-closing action and / or a preset lock-closing success prompt appears, determine that the bicycle is a shared bicycle.

2. The method for feedback on the failure of a shared bicycle according to claim 1, wherein The preset lock-closing action includes an action of sliding to close the bicycle lock or an action of closing the bicycle lock using a mobile phone; the preset lock-closing success prompt includes a voice prompt of successful lock-closing.

3. The shared bicycle fault feedback method according to claim 2, wherein, When the preset lock-closing action includes an action of sliding to close the bicycle lock, the determining whether a preset lock-closing action appears includes: Identify the bicycle lock in the video stream and determine whether the lock is a preset lock; In response to the lock being a preset lock, identify the position of the lock and the position of the target joint point corresponding to the preset lock-closing action; Detect the first overlap rate between the target joint point and the lock; In response to the first overlap rate being greater than the first ratio and maintaining for a preset duration, determine that a preset lock-closing action appears.

4. The shared bicycle fault feedback method according to claim 3, characterized in that, The identifying the position of the lock and the position of the target joint point corresponding to the preset lock-closing action includes: Frame by frame, locate the position of the lock and the target joint point corresponding to the preset lock-closing action; Calculate the centroid position of the target joint point, and assign a unique ID to the centroid; Calculate the distance between the centroids in two adjacent frames, associate the centroid with the smallest distance in two adjacent frames, and update the ID of the centroid in the current frame with the ID of the centroid in the previous frame to associate the same centroid in two adjacent frames; The detecting the first overlap rate between the target joint point and the lock includes: Detect the first overlap rate between the centroid and the lock.

5. The method for feedback on the failure of a shared bicycle according to claim 4, wherein The identifying the position of the lock and the position of the target joint point corresponding to the preset lock-closing action further includes: If a new target joint point is located, calculate the centroid position of the new target joint point and assign a unique ID to the centroid; and / or If the target joint point corresponding to the centroid with an assigned ID cannot be located for a continuous set number of frames, delete the centroid and its ID.

6. The shared bicycle fault feedback method according to claim 1, characterized in that The identifying the shared bicycles parked in the target parking area based on the video stream includes: Detect whether a batch of bicycles of the same color are parked in the target parking area in the video stream; In response to a batch of bicycles of the same color being parked in the target parking area, randomly identify the locks of one or more bicycles in the batch; If the randomly identified lock is a preset lock, determine that the batch of bicycles are shared bicycles.

7. The method for feedback on shared bicycle failures according to claim 1, characterized in that, The preset fault scenario includes: The first failure scenario: A person performs a preset unlocking action on a shared bicycle in the target parking area, and after the preset unlocking action ends, the person moves while the shared bicycle does not move; The second failure scenario: A person parks a shared bicycle in the target parking area, and within a preset time after parking, performs a preset unlocking action on another shared bicycle and leaves the target parking area; The third failure scenario: Within a preset time range, the shared bicycle in the target parking area has not been moved within a set duration.

8. The shared bicycle fault feedback method according to claim 7, wherein, In the case of a preset failure scenario occurring, record the failure score of the shared bicycle based on a preset rule, including: For a shared bicycle that has the first failure scenario occur continuously for a preset number of times, record a first failure score once; For a shared bicycle that has the second failure scenario occur, record a second failure score once for each occurrence of the second failure scenario; For a shared bicycle that has the third failure scenario occur, record a third failure score once for each set duration that it has not been moved; Wherein, the first failure score > the second failure score > the third failure score.

9. The method for feedback on shared bicycle failures according to claim 7, characterized in that, The preset unlocking action in the first failure scenario or the second failure scenario includes the action of using a mobile phone to unlock the vehicle, and the preset unlocking action is detected in the following manner: Detect the second overlap rate between the target joint point corresponding to the preset unlocking action and the mobile phone; In response to the second overlap rate reaching a second ratio, locate the position of the vehicle lock and the target joint point corresponding to the preset unlocking action frame by frame; Calculate the centroid position of the target joint point, and assign a unique ID to the centroid; Calculate the distance between the centroids in two adjacent frames, associate the centroid with the smallest distance in two adjacent frames, and update the ID of the centroid in the current frame with the ID of the centroid in the previous frame to associate the same centroid in two adjacent frames; Detect the first overlap rate between the centroid and the vehicle lock; In response to the first overlap rate being greater than a first ratio and maintaining a preset duration, determine that a preset unlocking action has occurred.

10. The method for feedback on the failure of a shared bicycle according to claim 1, wherein Based on the failure scores of each shared bicycle in the target parking area, feedback maintenance information, including: Based on the cumulative failure scores of each shared bicycle in the target parking area, determine the maintenance level of each shared bicycle, wherein, the higher the cumulative failure score, the higher the corresponding maintenance level; When the number of shared bicycles in the target parking area whose maintenance level reaches a preset maintenance level reaches a preset number, feedback maintenance information; wherein, the maintenance information includes the information of shared bicycles with different maintenance levels in the target parking area.

11. The shared bicycle fault feedback method according to claim 1, wherein, Based on the failure scores of each shared bicycle in the target parking area, feedback maintenance information, including: Based on the cumulative failure scores of each shared bicycle in the target parking area, determine the maintenance level of each shared bicycle, wherein, the higher the cumulative failure score, the higher the corresponding maintenance level; Calculate the sum of the cumulative failure scores of each shared bicycle in the target parking area, When the sum of the cumulative failure scores reaches a preset score, feedback maintenance information; wherein, the maintenance information includes the information of shared bicycles with different maintenance levels in the target parking area.

12. The shared bicycle fault feedback method according to claim 1, wherein The method further includes: Collect the shared bicycle data of the target parking area and form a visualization report according to a preset period; Among them, the shared bicycle data includes at least one of the number of shared bicycles parked in the target parking area within a preset period, the parking time of the shared bicycles, and the gender of the personnel parking the shared bicycles.

13. The method for feedback on shared bicycle faults according to claim 12, wherein It also includes: According to the collected shared bicycle data, provide a suggestion on the optimal number of shared bicycles parked in the target parking area; and / or Collect the number of shared bicycles parked outside the target parking area, determine whether the demarcation of the target parking area is reasonable, and provide a demarcation suggestion for the target parking area when the demarcation of the target parking area is unreasonable.

14. The method for feedback on the failure of a shared bicycle according to claim 13, wherein The providing a suggestion on the optimal number of shared bicycles parked in the target parking area according to the collected shared bicycle data includes: Count the traffic flow of the target parking area; Based on the average value of the traffic flow, determine the optimal number of shared bicycles parked in the target parking area; Among them, the optimal number of shared bicycles parked includes the optimal number of shared bicycles parked on weekdays and / or the optimal number of shared bicycles parked on rest days.

15. The shared bicycle fault feedback method according to claim 12, characterized in that, The collecting the shared bicycle data of the target parking area and forming a visualization report according to a preset period includes: Collect the shared bicycle data of the target parking area within a preset period; Convert the shared bicycle data into a target data format; Perform missing value processing on the shared bicycle data in the target data format to complete the missing fields; Perform outlier processing on the shared bicycle data after missing value processing, and the outlier processing includes deleting duplicate data; Form a visualization report based on the shared bicycle data after outlier processing.

16. A shared bicycle fault feedback device, characterized in that, It includes: An acquisition module for acquiring a video stream of a target parking area; An identification module for identifying shared bicycles parked in the target parking area based on the video stream; The identifying shared bicycles parked in the target parking area based on the video stream includes: detecting whether a single person parks a bicycle in the target parking area in the video stream; in response to a single person parking a bicycle in the target parking area, determining whether a preset locking action and / or a preset locking success prompt appears; if a preset locking action and / or a preset locking success prompt appears, determining that the bicycle is a shared bicycle; A detection module for detecting whether a preset failure scenario occurs for the shared bicycles in the target parking area based on the video stream; A recording module for recording the failure score of the shared bicycle based on a preset rule when a preset failure scenario occurs; A feedback module for feeding back maintenance information based on the failure scores of the shared bicycles in the target parking area.

17. A storage medium, characterized in that, A computer program is stored on the storage medium, and when the computer program is executed by one or more processors, the shared bicycle failure feedback method described in any one of claims 1 to 15 is implemented.

18. An electronic device, characterized in that, It includes a memory and one or more processors, and a computer program is stored on the memory, and when the computer program is executed by the one or more processors, the shared bicycle failure feedback method described in any one of claims 1 to 15 is implemented.

Images