A cash clearing video clip acquisition method and device

Abstract

The application provides a cash clearing video clip acquisition method and device, relates to the technical field of data acquisition, and can be applied to the financial field or other technical fields.The method comprises the following steps: acquiring reported amounts and cash box information; determining a cash-loaded cash box reference weight interval according to a first preset correspondence relationship, the reported amounts and the cash; determining a target search data set for data query according to a comparison result of the cash box weight and the cash box reference weight interval; determining a corresponding video clip acquisition mode according to an identification result of the cash box serial number; and acquiring a cash clearing video clip corresponding to the cash box in the target search data set by using the video clip acquisition mode.The device executes the above method.The cash clearing video clip acquisition method and device provided in the application can improve the work efficiency of acquiring the cash clearing video clip in the cash box.

Description

Technical Field

[0001] This invention relates to the field of data acquisition technology, specifically to a method and apparatus for acquiring video clips of cash sorting. Background Technology

[0002] Currently, before commercial banks sort cash through sealed packages, staff in the area where the cash box is being sorted need to record video with the side of the cash box bearing the name facing the camera, and also record the cash sorting process throughout the sealing of the cash box. This video recording is used to quickly locate the relevant cash box's cash sorting video clip if there is a discrepancy between the sorted amount and the actual reported amount, thus allowing for the assignment of responsibility to the staff. However, in practice, staff may forget to record the side of the cash box bearing the name facing the camera, and even if they do, there may be issues with the markings not being recognized, making it time-consuming and laborious to find the corresponding cash sorting video clip later. Summary of the Invention

[0003] To address the problems in the prior art, embodiments of the present invention provide a method and apparatus for obtaining cash sorting video clips, which can at least partially solve the problems existing in the prior art.

[0004] On one hand, this invention proposes a method for obtaining video clips of cash sorting, including:

[0005] Obtain the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash was sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value;

[0006] Based on the first preset correspondence, the reported amount, and the cash, a baseline weight range for cash boxes containing cash is determined; wherein, the first preset correspondence includes the correspondence between the preset weight range of cash boxes containing preset cash types and the preset reported amount;

[0007] Based on the comparison results between the weight of the cash box and the baseline weight range of the cash box, a target retrieval dataset for data query is determined. Based on the identification results of the cash box serial number, a corresponding video segment acquisition method is determined. The video segment acquisition method is used to obtain the cash sorting video segment corresponding to the cash box from the target retrieval dataset.

[0008] The step of determining the target retrieval dataset for data querying based on the comparison results between the weight of the carton and the carton's baseline weight range includes:

[0009] If it is determined that the weight of the box is within the reference weight range of the box, then the target retrieval dataset is determined to be the first retrieval dataset with normal weight.

[0010] If it is determined that the weight of the box is outside the reference weight range of the box, then the target retrieval dataset is determined as a second retrieval dataset marked with weight anomalies.

[0011] The method for determining the corresponding video segment acquisition based on the identification result of the box serial number includes:

[0012] If the identification result is determined to be that the number of digits in the cash box number is equal to a preset value and does not include letters, then the video segment acquisition method is determined to be to determine the cash sorting video segment in the cash box corresponding to the cash box number according to the second preset correspondence.

[0013] The second preset correspondence includes the correspondence between the preset cash box number and the preset cash sorting video segment identifier.

[0014] The method for determining the corresponding video segment acquisition based on the identification result of the box serial number includes:

[0015] If it is determined that the identification result is that the number of digits of the cash box serial number is not equal to the preset value and / or includes letters, then the cash box cash sorting location is obtained according to the location name of the cash box cash sorting location, and a cash sorting video taken in the cash box cash sorting location is obtained.

[0016] Image recognition is performed on the cash sorting video to obtain frame images containing information marked on the external surface of the cash box;

[0017] Each frame image is compared with the image of the cash box containing information about the outer surface of the cash box. The target video segment corresponding to the target frame image with the highest similarity comparison result is taken as the cash sorting video segment in the cash box.

[0018] The cash box information also includes a passive tag number; correspondingly, the method for obtaining the cash sorting video clip also includes:

[0019] If the identification result is determined to be an error in the number of digits in the cash box serial number and / or the inclusion of letters, then the set of passive tag numbers for the cash box cash sorting area is obtained;

[0020] The cash sorting video segment corresponding to the passive tag number is determined according to the third preset correspondence relationship; wherein, the third preset correspondence relationship includes the correspondence between each preset passive tag number in the passive tag number set and the preset cash sorting video segment identifier.

[0021] The identification of the item box serial number includes:

[0022] The serial number of the box is extracted using a preset character recognition model to obtain the recognition result;

[0023] The preset text recognition model is obtained by labeling and training a machine learning model based on the features of the box image.

[0024] Prior to the step of obtaining the reported amount and cash box information, the method for obtaining the cash sorting video clip further includes:

[0025] Obtain the cleared amount. If it is determined that the cleared amount and the reported amount are inconsistent, then proceed with obtaining the reported amount and cash box information, as well as subsequent steps.

[0026] On one hand, the present invention proposes a device for acquiring video clips of cash sorting, comprising:

[0027] The first acquisition unit is used to acquire the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash was sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value;

[0028] The determining unit is configured to determine a baseline weight range for a cash box containing cash based on a first preset correspondence, the reported amount, and the cash; wherein the first preset correspondence includes a correspondence between a preset weight range for a cash box containing a preset type of cash and a preset reported amount;

[0029] The second acquisition unit is used to determine the target retrieval dataset for data query based on the comparison result between the weight of the cash box and the reference weight range of the cash box, determine the corresponding video segment acquisition method based on the identification result of the identification of the cash box serial number, and use the video segment acquisition method to acquire the cash sorting video segment corresponding to the cash box in the target retrieval dataset.

[0030] In another aspect, embodiments of the present invention provide an electronic device, including: a processor, a memory, and a bus, wherein,

[0031] The processor and the memory communicate with each other via the bus;

[0032] The memory stores program instructions that can be executed by the processor, and the processor can execute the following methods by calling the program instructions:

[0033] Obtain the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash was sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value;

[0034] Based on the first preset correspondence, the reported amount, and the cash, a baseline weight range for cash boxes containing cash is determined; wherein, the first preset correspondence includes the correspondence between the preset weight range of cash boxes containing preset cash types and the preset reported amount;

[0035] Based on the comparison results between the weight of the cash box and the baseline weight range of the cash box, a target retrieval dataset for data query is determined. Based on the identification results of the cash box serial number, a corresponding video segment acquisition method is determined. The video segment acquisition method is used to obtain the cash sorting video segment corresponding to the cash box from the target retrieval dataset.

[0036] This invention provides a non-transitory computer-readable storage medium, comprising:

[0037] The non-transitory computer-readable storage medium stores computer instructions that cause the computer to perform the following methods:

[0038] Obtain the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash was sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value;

[0039] Based on the first preset correspondence, the reported amount, and the cash, a baseline weight range for cash boxes containing cash is determined; wherein, the first preset correspondence includes the correspondence between the preset weight range of cash boxes containing preset cash types and the preset reported amount;

[0040] Based on the comparison results between the weight of the cash box and the baseline weight range of the cash box, a target retrieval dataset for data query is determined. Based on the identification results of the cash box serial number, a corresponding video segment acquisition method is determined. The video segment acquisition method is used to obtain the cash sorting video segment corresponding to the cash box from the target retrieval dataset.

[0041] The cash sorting video clip acquisition method and apparatus provided in this invention acquire reported amount and cash box information. The cash box information includes the cash box weight, the region name where the cash sorting is located marked on the outer surface of the cash box, and the corresponding cash box serial number. The cash in the cash box is of the same currency type and value. A baseline weight range for the cash box containing cash is determined based on a first preset correspondence, the reported amount, and the cash. The first preset correspondence includes a correspondence between a preset weight range for cash boxes containing a preset cash type and a preset reported amount. A target retrieval dataset for data query is determined based on the comparison result of the cash box weight and the baseline weight range. A corresponding video clip acquisition method is determined based on the recognition result of the cash box serial number. The cash sorting video clip corresponding to the cash box is acquired using the video clip acquisition method and from the target retrieval dataset, thereby improving the efficiency of acquiring cash sorting video clips from the cash box. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0043] Figure 1 This is a block diagram of the cash sorting auxiliary system.

[0044] Figure 2 This is a flowchart illustrating how to use the cash sorting auxiliary system.

[0045] Figure 3 This is a flowchart illustrating the cash sorting video clip acquisition method provided in an embodiment of the present invention.

[0046] Figure 4 This is a flowchart illustrating a method for obtaining video clips of cash sorting provided in another embodiment of the present invention.

[0047] Figure 5 This is a schematic diagram of the structure of a cash sorting video clip acquisition device provided in an embodiment of the present invention.

[0048] Figure 6 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments and descriptions of the present invention are used to explain the present invention, but are not intended to limit the present invention. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be arbitrarily combined with each other.

[0050] Before implementing the cash sorting video clip acquisition method of this invention, it is necessary to build a cash sorting auxiliary system composed of relevant equipment, such as... Figure 1 As shown, the cash sorting auxiliary system includes an auxiliary device 11, a back-end system 20, an edge computing device 16, a sorting machine 17, a sorting camera 18, and a video storage device 19;

[0051] The auxiliary device 11 includes a weight sensor 12, a passive tag reader 13, a tag-capturing camera 14, a speaker 15, and an MCU 25. The auxiliary device 11 is placed on the sorting machine 17.

[0052] The weight sensor 12 may include a scale for obtaining the weight of the cash box (containing cash) and sending the weight information to the edge computing device 16.

[0053] The passive tag reader 13 is used to identify the passive tag number of the cash box (inside the cash box) and send the passive tag number to the edge computing device 16.

[0054] The identification camera 14 is used to capture the region name and cash box number marked on the outer surface of the cash box, and send the image information containing the region name and cash box number to the edge computing device 16.

[0055] Speaker 15 is used to play a prompt sound when the cash in each box is sorted.

[0056] MCU 25 is the main control unit of auxiliary device 11, used to communicate with edge computing device 16.

[0057] The back-end system 20 includes a main control unit 21, a storage unit 22, a search unit 23, and an input unit 24.

[0058] The main control unit 21 is responsible for binding video recordings with various modal features, obtaining binding relationships (i.e., the first preset correspondence, the second preset correspondence, and the third preset correspondence), and overall scheduling and processing of search and display.

[0059] Storage unit 22 is used to store binding relationships, as well as information such as image information and weight.

[0060] Search unit 23 is used for staff to search video recording units, and can be searched based on information such as weight, images, and text.

[0061] Input unit 24 is used to input the amount of money to be settled into the system.

[0062] Cash sorting machine 17 is used to sort the cash in the cash box.

[0063] The cash sorting camera 18 is used to record video of the entire cash sorting process.

[0064] Video memory 19 is used to store video data captured by the clearing camera 18.

[0065] like Figure 2 The following is an explanation of how to use the cash sorting auxiliary system:

[0066] Step S101: The staff, i.e. the sorting personnel, begin to prepare for sorting and turn on the power switch of the auxiliary device.

[0067] Step S102: Place the cash box to be cleared on the auxiliary device and press the trigger button. At this time, the MCU obtains the trigger timestamp.

[0068] Step S103: The scale, passive tag reader, and tag camera begin to acquire images of the weight, passive tag number, and box, respectively.

[0069] Step S104: The MCU determines whether the weight, passive tag number, and image data have been acquired. If they have, proceed to S105; otherwise, proceed to S103.

[0070] Step S105: After acquiring the weight, passive tag number, and image data, the auxiliary device sends a completion command to the speaker, which is then played through the speaker for the staff to know. Simultaneously, the MCU sends the weight, passive tag number, image data, trigger timestamp, and other information to the edge computing device.

[0071] Step S106: After obtaining multimodal data, the edge computing device first processes the image information, using a text recognition model to extract characters. Then, it sends information such as timestamp, weight, passive tag number, and image data to the main control unit.

[0072] Step S107: After receiving the multimodal data sent by the edge computing device, the main control unit simultaneously receives the timestamp and amount information of the sorting amount entered by the input unit. It then combines and binds the multimodal data sent by the edge computing device and the data from the input unit, and stores it in the storage unit. Information such as weight, passive tag number, and image data serves as the recording index for the sorting operation of this box. The start and end times are used to extract corresponding segments from the video storage.

[0073] Figure 3 This is a flowchart illustrating a method for obtaining video clips of cash sorting according to an embodiment of the present invention, as shown below. Figure 3 As shown, the cash sorting video clip acquisition method provided in this embodiment of the invention includes:

[0074] Step S1: Obtain the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash is sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value.

[0075] Step S2: Determine the baseline weight range of the cash box containing cash based on the first preset correspondence, the reported amount, and the cash; wherein, the first preset correspondence includes the correspondence between the preset weight range of the cash box containing the preset cash type and the preset reported amount.

[0076] Step S3: Based on the comparison results between the weight of the cash box and the baseline weight range of the cash box, determine the target retrieval dataset for data query, determine the corresponding video segment acquisition method based on the identification result of the cash box serial number, and use the video segment acquisition method to obtain the cash sorting video segment corresponding to the cash box from the target retrieval dataset.

[0077] In step S1 above, the device acquires the reported amount and cash box information; the cash box information includes the weight of the cash box, the region name where the cash sorting is located marked on the outer surface of the cash box, and the corresponding cash box serial number; wherein, the cash in the cash box is of the same currency type and value. The device can be a computer device that performs this method, for example, it may include an edge computing device. It should be noted that the acquisition and analysis of data involved in this embodiment of the invention are authorized by the user. The reported amount can be understood as the cash amount reported by the customer for cash sorting, which is of the same currency type and value, as explained below:

[0078] If the currency type is banknotes and the denomination is 1 yuan, then the cash in the cash box is all 1 yuan banknotes; if the currency type is banknotes and the denomination is 100 yuan, then the cash in the cash box is all 100 yuan banknotes; if the currency type is coins and the denomination is 1 yuan, then the cash in the cash box is all 1 yuan coins.

[0079] The purpose of marking the area where cash is sorted and the corresponding cash box number on the outside of the cash box is to enable the camera to take pictures of the markings.

[0080] Since cash can be sorted in multiple regions, and each region has multiple corresponding cash box numbers, a specific cash box in a particular region can only be uniquely identified by the name of the region and the corresponding cash box number.

[0081] Prior to the step of obtaining the reported amount and cash box information, the method for obtaining the cash sorting video clip further includes:

[0082] If the amount being cleared and the amount reported are inconsistent, then proceed with obtaining the reported amount and cash box information, and subsequent steps. The cleared amount refers to the actual amount of cash counted by the bank. It is understood that if the cleared amount and the reported amount are consistent, then there is no need to execute the subsequent steps of this method.

[0083] If the amount cleared and the amount reported are inconsistent, then proceed with the steps described above to obtain the reported amount and cash box information, as well as subsequent steps.

[0084] In step S2 above, the device determines the baseline weight range of the cash box containing cash based on the first preset correspondence, the reported amount, and the cash; wherein, the first preset correspondence includes the correspondence between the preset weight range of the cash box containing the preset cash type and the preset reported amount. An example of the first preset correspondence is illustrated below:

[0085] If the banknote is 1 yuan, the preset weight range corresponding to the preset reported amount of 100,000 yuan is 10-12 jin; if the banknote is 100 yuan, the preset weight range corresponding to the preset reported amount of 100,000 yuan is 7-9 jin; if the coin is 1 yuan, the preset weight range corresponding to the preset reported amount of 100,000 yuan is 700-800 jin.

[0086] If the reported amount is 50,000 yuan and the cash is in 1-yuan banknotes, then the base weight range for the cash box is determined to be 5-6 jin (the weight of the cash box itself is not considered at this time).

[0087] In step S3 above, the device determines the target retrieval dataset for data query based on the comparison result between the weight of the cash box and the reference weight range of the cash box, determines the corresponding video segment acquisition method based on the recognition result of the cash box serial number, and uses the video segment acquisition method to obtain the cash sorting video segment corresponding to the cash box from the target retrieval dataset. The step of determining the target retrieval dataset for data query based on the comparison result between the weight of the cash box and the reference weight range of the cash box includes:

[0088] If the weight of the box is determined to be within the reference weight range of the box, then the target retrieval dataset is determined to be the first retrieval dataset marked as having normal weight; the first retrieval dataset refers to the dataset with normal weight.

[0089] If it is determined that the weight of the cash box is outside the baseline weight range of the cash box, then the target retrieval dataset is determined as the second retrieval dataset marked with weight anomalies. The second retrieval dataset refers to the dataset with weight anomalies. By dividing the database into the first retrieval dataset and the second retrieval dataset, subsequent data queries (i.e., using the video clip acquisition method to obtain the cash sorting video clip corresponding to the cash box from the target retrieval dataset) can be performed through either the first retrieval dataset or the second retrieval dataset, thereby reducing the amount of data that needs to be traversed during data queries and improving data query efficiency.

[0090] The method for determining the corresponding video segment acquisition based on the identification result of the box serial number includes:

[0091] If the identification result is determined to be that the number of digits in the cash box serial number is equal to the preset value and does not include letters, then the video segment acquisition method is determined to be the cash sorting video segment in the cash box corresponding to the cash box serial number based on the second preset correspondence. The preset value can be set independently according to the actual situation and can be selected as a 7-digit value. It should be noted that the correct cash box serial number is a number string composed of 7 digits, such as 1020204. If the cash box label is worn out after long-term use, it will cause data loss, such as 6 digits remaining; or a part of a certain value is missing and is incorrectly identified as a letter. For example, if the upper or lower half of the number "8" is worn away, it is easy to be incorrectly identified as the letter "o".

[0092] Therefore, if the cash box number is a string of 7 digits, it can be considered that the cash box number is correct, and the video clip of cash sorting in the cash box can be obtained directly using the second preset correspondence.

[0093] The second preset correspondence includes the correspondence between preset cash box serial numbers and preset cash sorting video segment identifiers. An example of the second preset correspondence is given below:

[0094] The preset cash box number 1020204 corresponds to the preset cash sorting video segment identifier C1020204. That is, C1020204 records the start and end times of the video segment. It can be understood that the video content between the start and end times is the video footage of the entire cash sorting process of cash box number 1020204.

[0095] Similarly, the preset cash box number 1020205 corresponds to the preset cash sorting video segment identifier C1020205.

[0096] The method for determining the corresponding video segment acquisition based on the identification result of the box serial number includes:

[0097] If the identification result indicates that the number of digits in the cash box serial number is not equal to a preset value and / or includes letters, then the region where the cash box is sorted is obtained based on the region name where the cash sorting takes place, and the cash sorting video taken in that region is retrieved. Referring to the above explanation, if the cash box serial number is not a 7-digit number string, it means that the corresponding cash sorting video segment cannot be accurately obtained from the cash box serial number. Since each cash box's cash sorting region has its own corresponding cash sorting video (composed of the sequential order of each cash sorting video segment), the cash sorting video taken in that region can be uniquely identified by the region name of the cash box.

[0098] Image recognition is performed on the cash sorting video to obtain frame images containing information about the external surface of the cash boxes. Existing mature technologies can be used to perform image recognition on the cash sorting video to obtain frame images containing information about the external surface of the cash boxes, including the region name where the cash sorting takes place and the corresponding cash box number. Each frame image corresponds to the region name where the cash sorting takes place and the corresponding cash box number for one cash box.

[0099] Each frame image is compared with a cash box image containing information about the outer surface of the cash box. The target video segment corresponding to the target frame image with the highest similarity comparison value is taken as the cash sorting video segment from the cash box. The cash box image refers to the image of the cash box from which the cash sorting video segment is to be obtained. The highest similarity comparison value indicates that the image frame is most similar to the aforementioned cash box image in terms of image content.

[0100] If the frame image corresponding to the above cash box 1020204 is the target frame image, then the target video segment is the video segment between the start time and end time of the cash box playback, thereby obtaining the cash sorting video segment in the cash box.

[0101] The cash box information also includes a passive tag number; correspondingly, the method for obtaining the cash sorting video clip also includes:

[0102] If the identification result is determined to be an error in the number of digits in the cash box serial number and / or the inclusion of letters, then the set of passive tag numbers for the cash box cash sorting area is obtained. Since each cash box cash sorting area can define its own passive tag number, the passive tag number needs to correspond to the cash box cash sorting area number. For example, the passive tag number set for cash box cash sorting area A is 0001-0100, and the passive tag number set for cash box cash sorting area B is 0001-0095. Even if the passive tag numbers in A and B are the same, they do not correspond to the same cash box. The following explanation uses the passive tag number set of A (0001-0100) as an example.

[0103] The cash sorting video segment corresponding to the passive tag number is determined according to a third preset correspondence; wherein, the third preset correspondence includes the correspondence between each preset passive tag number in the passive tag number set and a preset cash sorting video segment identifier. An example of the third preset correspondence is illustrated below:

[0104] The preset passive tag number 0001 corresponds to the preset cash sorting video segment C0001, the preset passive tag number 0002 corresponds to the preset cash sorting video segment C0002, and so on. If the passive tag number is 0080, then according to the correspondence between each preset passive tag number in the set of passive tag numbers 0001-0100 and the preset cash sorting video segment identifier, the cash sorting video segment corresponding to the passive tag number 0080 is determined to be C0080.

[0105] Identifying the item box serial number includes:

[0106] The serial number of the box is extracted using a preset text recognition model to obtain the recognition result; the serial number of the box can be input into the preset text recognition model and the output result of the preset text recognition model can be used as the character extraction and recognition result, for example, 1020204.

[0107] The preset text recognition model is obtained by labeling and training a machine learning model based on the features of the box image. The sample data for model training is obtained by labeling the box image features. The training method for training the machine learning model using the sample data can be any existing conventional model training method. The machine learning model can include neural networks and random forests, etc., without specific limitations.

[0108] like Figure 4 As shown in the embodiment of the present invention, the method for obtaining cash sorting video clips is briefly described as follows:

[0109] Step S201: When it is determined that the amount to be cleared and the amount to be reported are inconsistent, initiate a video clip search task and open the search interface of the search unit.

[0110] Step S202: Obtain the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash was sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value.

[0111] Step S203: If the weight of the carton is within the standard weight range, the weight of the carton is normal; if the weight of the carton is outside the standard weight range, the weight of the carton is abnormal.

[0112] Step S204: Determine the target retrieval dataset as the first retrieval dataset with normal labeled weight.

[0113] Step S205: Identify the box serial number.

[0114] Step S206: If the recognition result shows that the number of digits in the box number is equal to the preset value and does not include letters, then the box number is normal; if the recognition result shows that the number of digits in the box number is not equal to the preset value and / or includes letters, then the box number is abnormal.

[0115] Step S207: Determine the cash sorting video segment in the cash box corresponding to the cash box number according to the second preset correspondence.

[0116] Step S208: Determine the target retrieval dataset as the second retrieval dataset with labeled weight anomalies.

[0117] Step S209: Obtain the location of the cash sorting in the cash box based on the location name of the cash sorting in the cash box, and obtain the cash sorting video taken in the location of the cash sorting in the cash box.

[0118] Step S210: Perform image recognition on the cash sorting video to obtain a frame image containing information about the outer surface of the cash box.

[0119] Step S211: Compare the similarity of each frame image with the cash box image containing the information of the outer surface of the cash box, and take the target video segment corresponding to the target frame image with the largest similarity comparison result as the cash sorting video segment in the cash box.

[0120] Step S212: Obtain the set of passive tag numbers for the cash box cash sorting area.

[0121] Step S213: Determine the cash sorting video segment corresponding to the passive tag number according to the third preset correspondence.

[0122] Step S214: Obtain a video clip of the cash being sorted from the cash box.

[0123] The cash sorting video clip acquisition method provided in this invention comprehensively applies multimodal technologies such as weighbridges, passive IoT tags, and machine vision to solve compliance issues for sorting personnel and to quickly locate corresponding video clips when anomalies are detected, thereby improving work efficiency. A device is installed on the sorting platform, large enough to hold the cash box. After the box has been in place for a certain period, it can acquire multimodal features such as the weight of the package to be sorted, the passive tag number, and text markings on the box. Through the correlation retrieval of these multimodal features, the efficiency and accuracy of video clip retrieval are improved.

[0124] The cash sorting video clip acquisition method provided in this invention acquires the reported amount and cash box information. The cash box information includes the cash box weight, the region name where the cash sorting is located marked on the outer surface of the cash box, and the corresponding cash box serial number. The cash in the cash box is of the same currency type and value. A baseline weight range for the cash box containing cash is determined based on a first preset correspondence, the reported amount, and the cash. The first preset correspondence includes a correspondence between a preset weight range for cash boxes containing a preset cash type and a preset reported amount. A target retrieval dataset for data query is determined based on the comparison result of the cash box weight and the baseline weight range. A corresponding video clip acquisition method is determined based on the recognition result of the cash box serial number. The cash sorting video clip corresponding to the cash box is acquired using the video clip acquisition method from the target retrieval dataset, thereby improving the efficiency of acquiring cash sorting video clips from the cash box.

[0125] Further, determining the target retrieval dataset for data querying based on the comparison results between the weight of the carton and the carton's baseline weight range includes:

[0126] If the weight of the box is determined to be within the reference weight range of the box, then the target retrieval dataset is determined to be the first retrieval dataset with normal weight; the above description is provided and will not be repeated here.

[0127] If it is determined that the weight of the box is outside the baseline weight range for the box, then the target retrieval dataset is determined to be a second retrieval dataset marked with weight anomalies. This can be referred to the above explanation and will not be repeated here.

[0128] The cash sorting video clip acquisition method provided in this embodiment of the invention can improve data query efficiency.

[0129] Furthermore, the step of determining the corresponding video segment acquisition method based on the recognition result of identifying the item box serial number includes:

[0130] If the identification result is determined to be that the number of digits in the cash box number is equal to a preset value and does not include letters, then the method for obtaining the video segment is determined to be to determine the cash sorting video segment in the cash box corresponding to the cash box number according to the second preset correspondence; the above description can be referred to, and will not be repeated here.

[0131] The second preset correspondence includes the correspondence between the preset cash box serial number and the preset cash sorting video segment identifier. Refer to the above explanation; further details are omitted here.

[0132] The cash sorting video clip acquisition method provided in this embodiment of the invention can acquire cash sorting video clips from cash boxes more quickly. Refer to the above description; further details are omitted.

[0133] Furthermore, the step of determining the corresponding video segment acquisition method based on the recognition result of identifying the item box serial number includes:

[0134] If the identification result is determined to be that the number of digits in the cash box serial number is not equal to the preset value and / or includes letters, then the cash box cash sorting location is obtained based on the location name of the cash box cash sorting location, and a cash sorting video taken in the cash box cash sorting location is obtained; the above description can be referred to, and will not be repeated.

[0135] Image recognition is performed on the cash sorting video to obtain frame images containing information about the external surface of the cash box; this can be referred to the above description and will not be repeated here.

[0136] Each frame of the image is compared with the image of the cash box containing information about its external surface. The target video segment corresponding to the target frame with the highest similarity score is taken as the cash sorting video segment in the cash box. This is similar to the above explanation and will not be repeated here.

[0137] The cash sorting video clip acquisition method provided in this embodiment of the invention can further improve the efficiency of acquiring cash sorting video clips from cash boxes when the cash box serial number cannot be used.

[0138] Furthermore, the cash box information also includes a passive tag number; correspondingly, the method for obtaining the cash sorting video clip also includes:

[0139] If the identification result is determined to be an error in the number of digits in the cash box serial number and / or the inclusion of letters, then the set of passive tag numbers for the cash box cash sorting area is obtained; this can be referred to the above description and will not be repeated here.

[0140] The cash sorting video segment corresponding to the passive tag number is determined according to the third preset correspondence relationship; wherein, the third preset correspondence relationship includes the correspondence between each preset passive tag number in the passive tag number set and the preset cash sorting video segment identifier. Refer to the above description; further details are omitted.

[0141] The cash sorting video clip acquisition method provided in this embodiment of the invention can further improve the efficiency of acquiring cash sorting video clips from cash boxes when the cash box serial number cannot be used.

[0142] Further, the identification of the box serial number includes:

[0143] The serial number of the box is extracted using a preset character recognition model to obtain the recognition result; the above description is provided and will not be repeated here.

[0144] The preset text recognition model is obtained by labeling and training a machine learning model based on the features of the box image. This can be referred to the above description and will not be repeated here.

[0145] The cash sorting video clip acquisition method provided in this embodiment of the invention can improve text recognition efficiency.

[0146] Furthermore, prior to the step of obtaining the reported amount and cash box information, the method for obtaining the cash sorting video clip also includes:

[0147] If the amount to be cleared is determined to be inconsistent with the amount reported, then proceed with obtaining the reported amount and cash box information, as well as subsequent steps. Refer to the above explanation; further details are omitted here.

[0148] The cash sorting video clip acquisition method provided in this embodiment of the invention does not require further execution of the technical solution when the sorted amount and the reported amount are consistent, thereby saving data processing resources.

[0149] It should be noted that the cash sorting video clip acquisition method provided in this embodiment of the invention can be used in the financial field, or in any technical field other than the financial field. This embodiment of the invention does not limit the application field of the cash sorting video clip acquisition method.

[0150] Figure 5 This is a schematic diagram of the structure of a cash sorting video clip acquisition device provided in an embodiment of the present invention, as shown below. Figure 5 As shown, the cash sorting video clip acquisition device provided in this embodiment of the invention includes a first acquisition unit 501, a determination unit 502, and a second acquisition unit 503, wherein:

[0151] The first acquisition unit 501 is used to acquire the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash sorting is located marked on the outer surface of the cash box, and the corresponding cash box serial number; wherein, the cash in the cash box is of the same currency type and value; the determination unit 502 is used to determine the benchmark weight range of the cash box containing cash according to the first preset correspondence, the reported amount, and the cash; wherein, the first preset correspondence includes the correspondence between the preset weight range of the cash box containing the preset cash type and the preset reported amount; the second acquisition unit 503 is used to determine the target retrieval dataset for data query based on the comparison result of the cash box weight and the benchmark weight range of the cash box, determine the corresponding video segment acquisition method based on the recognition result of the cash box serial number, and use the video segment acquisition method to acquire the cash sorting video segment corresponding to the cash box in the target retrieval dataset.

[0152] Specifically, the first acquisition unit 501 in the device is used to acquire the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash sorting is located marked on the outer surface of the cash box, and the corresponding cash box serial number; wherein, the cash in the cash box is of the same currency type and value; the determination unit 502 is used to determine the benchmark weight range of the cash box containing cash according to the first preset correspondence, the reported amount, and the cash; wherein, the first preset correspondence includes the correspondence between the preset weight range of the cash box containing the preset cash type and the preset reported amount; the second acquisition unit 503 is used to determine the target retrieval dataset for data query based on the comparison result of the cash box weight and the benchmark weight range of the cash box, determine the corresponding video segment acquisition method based on the recognition result of the cash box serial number, and use the video segment acquisition method to acquire the cash sorting video segment corresponding to the cash box in the target retrieval dataset.

[0153] The cash sorting video clip acquisition device provided in this embodiment of the invention acquires reported amount and cash box information; the cash box information includes cash box weight, the region name where the cash sorting is located marked on the outer surface of the cash box, and the corresponding cash box serial number; wherein the cash in the cash box is of the same currency type and value; a baseline weight range for cash boxes containing cash is determined according to a first preset correspondence, the reported amount, and the cash; wherein the first preset correspondence includes the correspondence between a preset weight range for cash boxes containing a preset cash type and a preset reported amount; a target retrieval dataset for data query is determined based on the comparison result of the cash box weight and the baseline weight range of the cash box; a corresponding video clip acquisition method is determined based on the recognition result of the cash box serial number; and the cash sorting video clip corresponding to the cash box is acquired using the video clip acquisition method in the target retrieval dataset, thereby improving the efficiency of acquiring cash sorting video clips from cash boxes.

[0154] Furthermore, the determining unit 502 is specifically used for:

[0155] If it is determined that the weight of the box is within the reference weight range of the box, then the target retrieval dataset is determined to be the first retrieval dataset with normal weight.

[0156] If it is determined that the weight of the box is outside the reference weight range of the box, then the target retrieval dataset is determined as a second retrieval dataset marked with weight anomalies.

[0157] The cash sorting video clip acquisition device provided in this embodiment of the invention can improve data query efficiency.

[0158] Furthermore, the second acquisition unit 503 is specifically used for:

[0159] If the identification result is determined to be that the number of digits in the cash box number is equal to a preset value and does not include letters, then the video segment acquisition method is determined to be to determine the cash sorting video segment in the cash box corresponding to the cash box number according to the second preset correspondence.

[0160] The second preset correspondence includes the correspondence between the preset cash box number and the preset cash sorting video segment identifier.

[0161] The cash sorting video clip acquisition device provided in this embodiment of the invention can acquire cash sorting video clips from cash boxes more quickly.

[0162] Furthermore, the second acquisition unit 503 is specifically used for:

[0163] If it is determined that the identification result is that the number of digits of the cash box serial number is not equal to the preset value and / or includes letters, then the cash box cash sorting location is obtained according to the location name of the cash box cash sorting location, and a cash sorting video taken in the cash box cash sorting location is obtained.

[0164] Image recognition is performed on the cash sorting video to obtain frame images containing information marked on the external surface of the cash box;

[0165] Each frame image is compared with the image of the cash box containing information about the outer surface of the cash box. The target video segment corresponding to the target frame image with the highest similarity comparison result is taken as the cash sorting video segment in the cash box.

[0166] The cash sorting video clip acquisition device provided in this embodiment of the invention can further improve the efficiency of acquiring cash sorting video clips from cash boxes when the cash box serial number cannot be used.

[0167] Furthermore, the cash box information also includes a passive tag number; correspondingly, the cash sorting video clip acquisition device is also used for:

[0168] If the identification result is determined to be an error in the number of digits in the cash box serial number and / or the inclusion of letters, then the set of passive tag numbers for the cash box cash sorting area is obtained;

[0169] The cash sorting video segment corresponding to the passive tag number is determined according to the third preset correspondence relationship; wherein, the third preset correspondence relationship includes the correspondence between each preset passive tag number in the passive tag number set and the preset cash sorting video segment identifier.

[0170] The cash sorting video clip acquisition device provided in this embodiment of the invention can further improve the efficiency of acquiring cash sorting video clips from cash boxes when the cash box serial number cannot be used.

[0171] Furthermore, the second acquisition unit 503 is specifically used for:

[0172] The serial number of the box is extracted using a preset character recognition model to obtain the recognition result;

[0173] The preset text recognition model is obtained by labeling and training a machine learning model based on the features of the box image.

[0174] The cash sorting video clip acquisition device provided in this embodiment of the invention can improve text recognition efficiency.

[0175] Furthermore, prior to the step of obtaining the reported amount and cash box information, the cash sorting video clip acquisition device is also used for:

[0176] Obtain the cleared amount. If it is determined that the cleared amount and the reported amount are inconsistent, then proceed with obtaining the reported amount and cash box information, as well as subsequent steps.

[0177] The cash sorting video clip acquisition device provided in this embodiment of the invention does not need to continue executing the technical solution when the sorted amount and the reported amount are consistent, thereby saving data processing resources.

[0178] The embodiments of the present invention provide a cash sorting video clip acquisition device that can be used to execute the processing flow of the above method embodiments. Its functions will not be repeated here, but can be referred to the detailed description of the above method embodiments.

[0179] Figure 6 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as... Figure 6 As shown, the electronic device includes: a processor 601, a memory 602, and a bus 603;

[0180] The processor 601 and the memory 602 communicate with each other via the bus 603.

[0181] The processor 601 is used to call program instructions in the memory 602 to execute the methods provided in the above-described method embodiments, including, for example:

[0182] Obtain the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash was sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value;

[0183] Based on the first preset correspondence, the reported amount, and the cash, a baseline weight range for cash boxes containing cash is determined; wherein, the first preset correspondence includes the correspondence between the preset weight range of cash boxes containing preset cash types and the preset reported amount;

[0184] Based on the comparison results between the weight of the cash box and the baseline weight range of the cash box, a target retrieval dataset for data query is determined. Based on the identification results of the cash box serial number, a corresponding video segment acquisition method is determined. The video segment acquisition method is used to obtain the cash sorting video segment corresponding to the cash box from the target retrieval dataset.

[0185] This embodiment discloses a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, and when the program instructions are executed by a computer, the computer can perform the methods provided in the above-described method embodiments, such as:

[0186] Obtain the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash was sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value;

[0187] Based on the first preset correspondence, the reported amount, and the cash, a baseline weight range for cash boxes containing cash is determined; wherein, the first preset correspondence includes the correspondence between the preset weight range of cash boxes containing preset cash types and the preset reported amount;

[0188] Based on the comparison results between the weight of the cash box and the baseline weight range of the cash box, a target retrieval dataset for data query is determined. Based on the identification results of the cash box serial number, a corresponding video segment acquisition method is determined. The video segment acquisition method is used to obtain the cash sorting video segment corresponding to the cash box from the target retrieval dataset.

[0189] This embodiment provides a computer-readable storage medium storing a computer program that causes the computer to execute the methods provided in the above-described method embodiments, including, for example:

[0190] Obtain the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash was sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value;

[0191] Based on the first preset correspondence, the reported amount, and the cash, a baseline weight range for cash boxes containing cash is determined; wherein, the first preset correspondence includes the correspondence between the preset weight range of cash boxes containing preset cash types and the preset reported amount;

[0192] Based on the comparison results between the weight of the cash box and the baseline weight range of the cash box, a target retrieval dataset for data query is determined. Based on the identification results of the cash box serial number, a corresponding video segment acquisition method is determined. The video segment acquisition method is used to obtain the cash sorting video segment corresponding to the cash box from the target retrieval dataset.

[0193] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0194] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0195] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0196] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0197] In the description of this specification, the references to terms such as "an embodiment," "a specific embodiment," "some embodiments," "for example," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0198] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for obtaining video clips of cash sorting, characterized in that, include: Obtain the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash was sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value; Based on the first preset correspondence, the reported amount, and the cash, a baseline weight range for cash boxes containing cash is determined; wherein, the first preset correspondence includes the correspondence between the preset weight range of cash boxes containing preset cash types and the preset reported amount; Based on the comparison results between the weight of the cash box and the baseline weight range of the cash box, a target retrieval dataset for data query is determined. Based on the identification results of the cash box serial number, a corresponding video segment acquisition method is determined. The video segment acquisition method is used to obtain the cash sorting video segment corresponding to the cash box from the target retrieval dataset.

2. The method for obtaining cash sorting video clips according to claim 1, characterized in that, The step of determining the target retrieval dataset for data querying based on the comparison results between the weight of the carton and the carton's baseline weight range includes: If it is determined that the weight of the box is within the reference weight range of the box, then the target retrieval dataset is determined to be the first retrieval dataset with normal weight. If it is determined that the weight of the box is outside the reference weight range of the box, then the target retrieval dataset is determined as a second retrieval dataset marked with weight anomalies.

3. The method for obtaining cash sorting video clips according to claim 1, characterized in that, The method for determining the corresponding video segment acquisition based on the identification result of the box serial number includes: If the identification result is determined to be that the number of digits in the cash box number is equal to a preset value and does not include letters, then the video segment acquisition method is determined to be to determine the cash sorting video segment in the cash box corresponding to the cash box number according to the second preset correspondence. The second preset correspondence includes the correspondence between the preset cash box number and the preset cash sorting video segment identifier.

4. The method for obtaining cash sorting video clips according to claim 1, characterized in that, The method for determining the corresponding video segment acquisition based on the identification result of the box serial number includes: If it is determined that the identification result is that the number of digits of the cash box serial number is not equal to the preset value and / or includes letters, then the cash box cash sorting location is obtained according to the location name of the cash box cash sorting location, and a cash sorting video taken in the cash box cash sorting location is obtained. Image recognition is performed on the cash sorting video to obtain frame images containing information marked on the external surface of the cash box; Each frame image is compared with the image of the cash box containing information about the outer surface of the cash box. The target video segment corresponding to the target frame image with the highest similarity comparison result is taken as the cash sorting video segment in the cash box.

5. The method for obtaining cash sorting video clips according to claim 4, characterized in that, The cash box information also includes a passive tag number; correspondingly, the method for obtaining the cash sorting video clip also includes: If the identification result is determined to be an error in the number of digits in the cash box serial number and / or the inclusion of letters, then the set of passive tag numbers for the cash box cash sorting area is obtained; The cash sorting video segment corresponding to the passive tag number is determined according to the third preset correspondence relationship; wherein, the third preset correspondence relationship includes the correspondence between each preset passive tag number in the passive tag number set and the preset cash sorting video segment identifier.

6. The method for obtaining cash sorting video clips according to any one of claims 1 to 5, characterized in that, Identifying the item box serial number includes: The serial number of the box is extracted using a preset character recognition model to obtain the recognition result; The preset text recognition model is obtained by labeling and training a machine learning model based on the features of the box image.

7. The method for obtaining cash sorting video clips according to any one of claims 1 to 5, characterized in that, Prior to the step of obtaining the reported amount and cash box information, the method for obtaining the cash sorting video clip further includes: Obtain the cleared amount. If it is determined that the cleared amount and the reported amount are inconsistent, then proceed with obtaining the reported amount and cash box information, as well as subsequent steps.

8. A device for acquiring video clips of cash sorting, characterized in that, include: The first acquisition unit is used to acquire the reported amount and cash box information; the cash box information includes the weight of the cash box, the name of the region where the cash was sorted and the corresponding cash box serial number marked on the outer surface of the cash box; wherein, the cash in the cash box is of the same currency type and value; The determining unit is configured to determine a baseline weight range for a cash box containing cash based on a first preset correspondence, the reported amount, and the cash; wherein the first preset correspondence includes a correspondence between a preset weight range for a cash box containing a preset type of cash and a preset reported amount; The second acquisition unit is used to determine the target retrieval dataset for data query based on the comparison result between the weight of the cash box and the reference weight range of the cash box, determine the corresponding video segment acquisition method based on the identification result of the identification of the cash box serial number, and use the video segment acquisition method to acquire the cash sorting video segment corresponding to the cash box in the target retrieval dataset.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images