Completely unsupervised luggage re-identification method based on dynamic queue depth clustering

Abstract

The invention provides a completely unsupervised luggage re-identification method based on dynamic queue depth clustering, and the method mainly comprises the steps: creating a luggage data set LReID-50tohsound without an identity label, carrying out the target detection through employing YOLOv5, and generating 5000 luggage images with different identities, and the total number is 5000. A local video acquisition environment is established, and luggage check-in images and video data of a sorting line are obtained. A DynQueueViT luggage re-identification model is deployed, a first-in first-out dynamic queue management feature is adopted, the video memory requirement is reduced, and pseudo tag noise is reduced by using time constraint. A luggage image is decomposed into global features, local features and mixed features by the model, and dynamic queue updating training is carried out. And re-identification features are calculated through the multi-layer visual conversion module and the local feature data enhancement module. And finally, judging whether the luggage belongs to the same identity or not through a distance formula. The model introduces time constraint and a positive and negative sample matching algorithm, improves the re-recognition accuracy, and evaluates the identity similarity of luggage images by calculating the cosine distance.

Description

Technical Field

[0001] The present invention relates to the field of computer vision, and in particular to a completely unsupervised baggage re-identification method based on dynamic queue depth clustering. Background Art

[0002] With the accelerated development of the domestic economy and the continuous improvement of people's living standards, more and more people choose to travel by plane. During the flight, it is usually the case that a passenger will carry multiple pieces of luggage, which leads to the fact that the amount of luggage that domestic airports need to handle is far greater than the number of passengers. Faced with such a large amount of luggage, domestic airports can no longer meet people's travel needs by relying solely on manual sorting of luggage. In this case, for airports, the importance of automated baggage sorting systems in the process of transporting passengers has become increasingly prominent. At present, mainstream airport baggage automated sorting systems use radio frequency identification technology (RFID) to sort luggage.

[0003] Radio frequency identification technology uses unique electromagnetic tags and electromagnetic readers to identify and transmit data for baggage. As can be seen from the figure, the baggage sorting process is divided into two main stages: check-in and conveyor belt tracking. In the check-in stage, airport staff will attach electromagnetic tags with unique identification codes to the luggage for subsequent tracking and identification. In the subsequent conveyor belt tracking stage, special electromagnetic reading and writing equipment is responsible for the identification and tracking of luggage. Specifically, after the luggage is labeled at the check-in counter and enters the conveyor belt, it will move quickly along the conveyor belt and pass through multiple electromagnetic reading and writing device nodes. In this process, each node can identify the electromagnetic tags attached to the past luggage and integrate the node information that the electromagnetic tags pass through to form the movement trajectory of the luggage, such as passing through node #1, node #3, node #4, etc.

[0004] Although RFID technology has been widely used in baggage tracking and identification, it still has some inherent limitations. For example, the equipment cost is high, the baggage tracking effect is limited by the number of readers, and electromagnetic interference may cause identification failure. These problems limit the further development and promotion of automated baggage sorting systems. At the same time, in recent years, computer vision technology has demonstrated excellent robustness, accuracy and real-time performance in tracking and monitoring pedestrians and vehicles, showing the potential to handle complex scenes. Therefore, in order to meet people's strong demand for efficient travel, it is necessary to upgrade the baggage tracking system based on RFID to a baggage tracking system using computer vision technology.

[0005] The computer vision baggage tracking system realizes the entire cross-camera baggage tracking process by identifying, tracking, and re-identifying the baggage in the captured image. Specifically, this process first involves identifying which objects are baggage from the camera image. Next, the system monitors the movement trajectory of these baggage in real time to complete the single-camera tracking task. The critical last step is that the system needs to confirm the exact identity of these baggage and match the baggage trajectories captured by different cameras to construct the complete path of the baggage. In the process of cross-camera tracking, baggage re-identification technology plays a role similar to that of electromagnetic tags. It identifies the identity of the baggage by analyzing the characteristics of the baggage in the camera image. This is the core link to complete the cross-camera baggage trajectory matching and build a complete baggage path.

[0006] Thanks to the rapid development of neural network technology in recent years, luggage recognition and tracking within a single camera have shown excellent performance. However, in terms of re-identification technology for luggage identity matching across cameras, current solutions still do not fully meet the needs of practical applications. Therefore, this paper focuses on the method of luggage re-identification.

[0007] In the specific scenario of baggage sorting lines, object re-identification technology faces new challenges. These challenges mainly come from the fact that the accuracy of re-identification is highly dependent on the quality of the images captured by the camera and whether the appearance features of the luggage remain consistent in different scenarios. The operating environment of the baggage re-identification task is more complex than the common pedestrian or vehicle re-identification scenarios, and complex environmental conditions often cause the appearance features of the same piece of luggage to change.

[0008] In a baggage sorting flow environment, cameras are installed at multiple key locations on the conveyor belt, where other auxiliary sorting equipment is often located nearby.

[0009] The following is a brief description of how auxiliary sorting equipment causes changes in the appearance and posture of baggage:

[0010] (1) Signal indicator light

[0011] There are multiple timed flashing signal indicators on the baggage sorting line. The rapid alternation of light and dark causes the appearance color of some metallic painted luggage to change rapidly.

[0012] (2) Conveyor belt

[0013] The conveyor belt of the baggage sorting line has ups and downs and many bends. The baggage often reverses during the forward movement, and the posture of the baggage before and after the reversal is often quite different.

[0014] (3) Mechanical swing arm

[0015] The baggage sorting line uses mechanical swing arms for diversion, and the mechanical swing arms will also change the posture of the luggage when changing the luggage conveying channel.

[0016] (4) Fence

[0017] The fences on the baggage sorting line will block the baggage from the camera's field of view.

[0018] In summary, since the same piece of luggage may present significantly different surface colors and moving postures under different camera fields of view, and the camera images may sometimes not fully cover the target luggage, the luggage re-identification task faces the challenge of accurately identifying a large number of target identities in such a complex and changing environment.

[0019] In addition, in airport scenarios with a daily passenger flow of about 100,000 people, the baggage videos taken from different angles on the sorting line every day provide rich data resources for baggage re-identification model training. However, the huge size and complexity of these data make it unfeasible to rely on traditional manual annotation methods. Therefore, fully unsupervised training algorithms have become a better choice for training baggage re-identification models using large-scale datasets. However, most existing fully unsupervised re-identification training algorithms are mainly trained and tested on small-scale datasets. In environments with small data volumes, these algorithms require a large amount of video memory resources for data comparison and clustering. This approach makes model training more difficult when processing large-scale datasets. Therefore, another challenge facing unsupervised training for baggage re-identification is how to handle the training of large-scale datasets with limited video memory resources.

[0020] In addition, this paper is based on the school-level project "Development of Key Technologies for Multi-source Data Fusion and Collaboration in Aviation Logistics" led by the Second Research Institute of the Civil Aviation Administration of China. In summary, in-depth research and development of baggage re-identification models is not only an important scientific research topic, but also has great economic significance. Summary of the invention

[0021] The purpose of this invention is to address the problem that the existing fully unsupervised re-identification model has too high video memory requirements during training, which makes it difficult to train the model on a large-scale unlabeled baggage dataset. This paper proposes a fully unsupervised baggage re-identification model DynQueueViT based on dynamic queue deep clustering. This model reduces the video memory requirements for training and improves the re-identification performance of the model.

[0022] The object of the present invention is achieved in this way:

[0023] A fully unsupervised baggage re-identification model based on dynamic queue deep clustering, DynQueueViT, is proposed. The model aims to make full use of the large amount of unlabeled data generated by the baggage sorting pipeline and improve the accuracy of baggage re-identification under the condition of limited graphics card resources.

[0024] The specific method is as follows:

[0025] 1. A completely unsupervised baggage re-identification method based on dynamic queue depth clustering, characterized in that the method comprises the following steps:

[0026] Step 1: Construct a baggage dataset and create a baggage dataset without identity tags, LReID-50tohsouand. These videos are recorded by five cameras, covering various locations of the sorting line, with a total duration of 148 hours, and contain more than 100,000 pieces of baggage with different identities. The dataset was created using the YOLOv5 tool. YOLOv5 is used for target detection, and the detection results are cropped to construct the dataset. LReID-50tohsouand contains 5,000 pieces of baggage with different identities, a total of 50,000 pictures, and an average of 10 pictures of each piece of baggage from different angles.

[0027] Step 2: Build a local baggage sorting line video acquisition environment, including multiple non-overlapping cameras and a video stream transmission server for the baggage sorting line;

[0028] Step 3: Build a local baggage check-in video acquisition environment, including a baggage check-in camera and a video stream transmission server;

[0029] Step 4: Obtain baggage check-in images and videos of baggage taken under different viewing areas on the water flow sorting line;

[0030] Step 5. Deploy the DynQueueViT model for local baggage re-identification feature calculation. The DynQueueViT baggage re-identification model does not store all the features of the training images in the video memory at once, but uses a first-in, first-out dynamic queue to manage these features. As the training progresses, the features in the queue will be continuously updated to ensure that there are enough samples to participate in the training while significantly reducing the demand for video memory for training. At the same time, during the training process, the DynQueueViT baggage re-identification model uses the time constraints of the video sequence for the training data entering the model, which reduces the noise of pseudo labels to a certain extent.

[0031] Step 6: Calculate the re-identification features between different luggage using the local luggage re-identification feature calculation model, using the multi-layer visual conversion module and the local feature data enhancement module;

[0032] Step 7: Substitute the re-identification features between different luggage into the distance formula; determine whether the two target luggages have the same identity;

[0033] 2. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that the baggage dataset LReID-50tohsouand without identity tags in step 1 is created. The production steps include: using LabelMe and video to produce a small batch detection dataset, initializing and fine-tuning the YOLOv5 network, generating a video with a baggage bounding box labeled, extracting a picture every 4 frames, cropping the bounding box to generate a spare dataset, and finally manually screening to exclude duplicate pictures, and completing the manual labeling of the unlabeled dataset;

[0034] 3. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that the DynQueueViT model in step 5 uses a feature extraction network to decompose the baggage image input to the model into global features f g , local features f l Compare feature f with weight reference data ;

[0035] 4. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that the DynQueueViT model in step 5 is a global feature f of the input image. g and local features f l A first-in-first-out dynamic queue is established for each. At the same time, in order to capture image information more comprehensively, a dynamic queue for storing mixed features is also introduced. When evaluating whether a sample pair belongs to a positive sample pair or a negative sample pair, the DynQueueViT baggage re-identification model mainly relies on global and local feature queues. When defining the contrast loss to train the feature extraction network, the DynQueueViT baggage re-identification model chooses to use mixed features.

[0036] 5. A completely unsupervised baggage re-identification method based on deep clustering of dynamic queues according to claim 1, characterized in that the DynQueueViT model in step 5 introduces time constraints of the video sequence when updating the dynamic queue. Specifically, in an actual baggage sorting line environment, multiple cameras will shoot the baggage at the same time, and the same piece of baggage is likely to be captured by different cameras in the same time period. Therefore, using the shooting timestamp attached to each image in the LRe-ID50thsouand dataset, images taken by different cameras in the same time period are classified into the same training batch, and then the training batch images are sent to the DynQueueViT baggage re-identification model for training.

[0037] 6. A completely unsupervised baggage re-identification method based on deep clustering of dynamic queues according to claim 1, characterized in that in step 5, the DynQueueViT model uses a positive and negative sample matching algorithm to find all baggage images belonging to the same identity in the dynamic queue and identifies them as a positive sample set. Correspondingly, the hybrid features obtained by fusing all image features in the positive sample set will be used as the feature representation of the anchor sample.

[0038] For all images in the dynamic queue {X i ,i=1,2…,M×batch_size}, and calculate their global features f respectively. g and local features f l With all images in the dynamic queue {X j ,j=1,2…,M×batch_size} (including the image itself). The image features in the dynamic queue are assigned to different similarity sets. The positive and negative sample matching algorithm will also assign the images in the queue to dissimilar sets and boundary sets. i The distance is far, located at the total distance S total (X i ,X j ) from the image at the end of ascending order, forming image X i The dissimilar set ImageX i The boundary set Then the excluded similar candidate sets Image X j With those in total distance S total (X i ,X j ) in ascending order but not in the candidate set image composition.

[0039] In selecting the positive sample set, from image X i Similarity set Select all samples whose distance is less than the threshold value to form a positive sample set In selecting the negative sample set, we take the image X i The boundary set Dissimilar Sets Some pictures in the middle constitute the negative sample set

[0040] 7. The ViT baggage re-identification method based on local data enhancement according to claim 1 is characterized in that the LocEnViT model in step 5 is obtained by fusing the positive sample set The global and local features of all images in create a comprehensive global feature f i mg and local features f i ml

[0041]

[0042] 8. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that the DynQueueViT model in step 5 uses a hybrid feature contrast loss L m and data enhancement contrast loss L data Train the model. The overall loss function L total It can be expressed as:

[0043] L total =βL m +(1-β)L data

[0044] Hybrid feature contrast loss L m The feature extraction network is trained by using a set of positive samples of real data. This loss is further divided into two parts: the global mixed feature L mg Contrastive loss and local mixed feature contrast loss L ml :

[0045]

[0046] The contrast loss can be expressed as:

[0047]

[0048] 9. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that in said step 5, when the DynQueueViT model determines whether two baggage images belong to the same identity, the LocEnViT baggage re-identification model performs cosine distance calculation:

[0049] BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic diagram of the unlabeled baggage dataset LReID-50thsouand.

[0051] Figure 2 It is a schematic diagram of the training framework of a fully unsupervised re-identification model.

[0052] Figure 3 It is a schematic diagram of the deep clustering framework based on dynamic queues.

[0053] Figure 4 This is a schematic diagram of the dynamic queue update process. DETAILED DESCRIPTION

[0054] The specific implementation modes of the present invention are further described in detail below with reference to the accompanying drawings.

[0055] like Figure 2 As shown, the present invention first constructs a baggage dataset: creates a baggage dataset LReID-50tohsouand without identity tags, uses YOLOv5 for target detection, and constructs a dataset by cropping the detection results. The dataset contains 5,000 pieces of baggage with different identities, a total of 50,000 pictures, and an average of 10 pictures of each piece of baggage from different angles. Local video acquisition environment construction: Construct a baggage sorting line and baggage check-in video acquisition environment, including multiple cameras and video stream transmission servers. Baggage image and video acquisition: Obtain baggage check-in images and videos shot under different viewing areas on the flow sorting line. DynQueueViT model deployment and training: Deploy the DynQueueViT baggage re-identification model, adopt the first-in-first-out dynamic queue management feature, reduce the demand for video memory, and use the time constraints of the video sequence to reduce pseudo-label noise. The model decomposes the input baggage image into global features, local features, and mixed features, and uses dynamic queue updates and time constraints for training. Re-identification feature calculation: Use a multi-layer visual conversion module and a local feature data enhancement module to calculate the re-identification features between different baggage. Identity judgment: Substitute the re-identification features between different luggage into the distance formula to determine whether the two target luggages belong to the same identity.

[0056] The DynQueueViT model improves the accuracy of re-identification by introducing time constraints and positive and negative sample matching algorithms. When defining the contrast loss, the global and local feature queues choose to use mixed features. The model is trained using mixed feature contrast loss and data enhancement contrast loss. Finally, the cosine distance is calculated to determine whether the two luggage images belong to the same identity, which includes the following steps:

[0057] 1. A completely unsupervised baggage re-identification method based on dynamic queue depth clustering, characterized in that the method comprises the following steps:

[0058] Step 1: Construct a baggage dataset and create a baggage dataset without identity tags, LReID-50tohsouand. These videos are recorded by five cameras, covering various locations of the sorting line, with a total duration of 148 hours, and contain more than 100,000 pieces of baggage with different identities. The dataset was created using the YOLOv5 tool. YOLOv5 is used for target detection, and the detection results are cropped to construct the dataset. LReID-50tohsouand contains 5,000 pieces of baggage with different identities, a total of 50,000 pictures, and an average of 10 pictures of each piece of baggage from different angles.

[0059] Step 2: Build a local baggage sorting line video acquisition environment, including multiple non-overlapping cameras and a video stream transmission server for the baggage sorting line;

[0060] Step 3: Build a local baggage check-in video acquisition environment, including a baggage check-in camera and a video stream transmission server;

[0061] Step 4: Obtain baggage check-in images and videos of baggage taken under different viewing areas on the water flow sorting line;

[0062] Step 5. Deploy the DynQueueViT model for local baggage re-identification feature calculation. The DynQueueViT baggage re-identification model does not store all the features of the training images in the video memory at once, but uses a first-in, first-out dynamic queue to manage these features. As the training progresses, the features in the queue will be continuously updated to ensure that there are enough samples to participate in the training while significantly reducing the demand for video memory for training. At the same time, during the training process, the DynQueueViT baggage re-identification model uses the time constraints of the video sequence for the training data entering the model, which reduces the noise of pseudo labels to a certain extent.

[0063] Step 6: Calculate the re-identification features between different luggage using the local luggage re-identification feature calculation model, using the multi-layer visual conversion module and the local feature data enhancement module;

[0064] Step 7: Substitute the re-identification features between different luggage into the distance formula; determine whether the two target luggages have the same identity;

[0065] 2. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that the baggage dataset LReID-50tohsouand without identity tags in step 1 is created. The production steps include: using LabelMe and video to produce a small batch detection dataset, initializing and fine-tuning the YOLOv5 network, generating a video with a baggage bounding box labeled, extracting a picture every 4 frames, cropping the bounding box to generate a spare dataset, and finally manually screening to exclude duplicate pictures, and completing the manual labeling of the unlabeled dataset;

[0066] 3. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that the DynQueueViT model in step 5 uses a feature extraction network to decompose the baggage image input to the model into global features f g , local features f l Compare feature f with weight reference data ;

[0067] 4. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that the DynQueueViT model in step 5 is a global feature f of the input image. g and local features f l A first-in-first-out dynamic queue is established for each. At the same time, in order to capture image information more comprehensively, a dynamic queue for storing mixed features is also introduced. When evaluating whether a sample pair belongs to a positive sample pair or a negative sample pair, the DynQueueViT baggage re-identification model mainly relies on global and local feature queues. When defining the contrast loss to train the feature extraction network, the DynQueueViT baggage re-identification model chooses to use mixed features.

[0068] 5. A completely unsupervised baggage re-identification method based on deep clustering of dynamic queues according to claim 1, characterized in that the DynQueueViT model in step 5 introduces time constraints of the video sequence when updating the dynamic queue. Specifically, in an actual baggage sorting line environment, multiple cameras will shoot the baggage at the same time, and the same piece of baggage is likely to be captured by different cameras in the same time period. Therefore, using the shooting timestamp attached to each image in the LRe-ID50thsouand dataset, images taken by different cameras in the same time period are classified into the same training batch, and then the training batch images are sent to the DynQueueViT baggage re-identification model for training.

[0069] 6. A completely unsupervised baggage re-identification method based on deep clustering of dynamic queues according to claim 1, characterized in that in step 5, the DynQueueViT model uses a positive and negative sample matching algorithm to find all baggage images belonging to the same identity in the dynamic queue and identifies them as a positive sample set. Correspondingly, the hybrid features obtained by fusing all image features in the positive sample set will be used as the feature representation of the anchor sample.

[0070] For all images in the dynamic queue {X i ,i=1,2…,M×batch_size}, and calculate their global features f respectively. g and local features f l With all images in the dynamic queue {X j ,j=1,2…,M×batch_size} (including the image itself). The image features in the dynamic queue are assigned to different similarity sets. The positive and negative sample matching algorithm will also assign the images in the queue to dissimilar sets and boundary sets. i The distance is far, located at the total distance S total (X i ,X j ) from the image at the end of ascending order, forming image X i The dissimilar set ImageX i The boundary set Then the excluded similar candidate sets Image X j With those in total distance S total (X i ,X j ) in ascending order but not in the candidate set image composition.

[0071] In selecting the positive sample set, from image X i Similarity set Select all samples whose distance is less than the threshold value to form a positive sample set In selecting the negative sample set, we take the image X i The boundary set Dissimilar Sets Some pictures in the middle constitute the negative sample set

[0072] 7. The ViT baggage re-identification method based on local data enhancement according to claim 1 is characterized in that the LocEnViT model in step 5 is obtained by fusing the positive sample set The global and local features of all images in create a comprehensive global feature fi mg and local features f i ml :

[0073] 8. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that the DynQueueViT model in step 5 uses a hybrid feature contrast loss L m and data enhancement contrast loss L data Train the model. The overall loss function L total It can be expressed as: total =βL m +(1-β)L data .

[0074] Hybrid feature contrast loss L m The feature extraction network is trained by using a set of positive samples of real data. This loss is further divided into two parts: the global mixed feature L mg Contrastive loss and local mixed feature contrast loss L ml :

[0075]

[0076]

[0077] The contrast loss can be expressed as:

[0078] 9. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that in said step 5, when the DynQueueViT model determines whether two baggage images belong to the same identity, the LocEnViT baggage re-identification model performs cosine distance calculation:

Claims

1. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering, characterized in that: The method comprises the following steps: Step 1: Construct a baggage dataset. Create a baggage dataset without identity tags, LReID-50tohsouand. These videos are recorded by five cameras, covering various locations of the sorting line, with a total duration of 148 hours, and contain more than 100,000 pieces of baggage with different identities. The dataset is made using the YOLOv5 tool. YOLOv5 is used for target detection, and the detection results are cropped to construct the dataset. LReID-50tohsouand contains 5,000 pieces of baggage with different identities, a total of 50,000 pictures, and an average of 10 pictures of each piece of baggage from different angles. Step 2: Build a local baggage sorting line video acquisition environment, including multiple non-overlapping cameras and a video stream transmission server for the baggage sorting line; Step 3: Build a local baggage check-in video acquisition environment, including a baggage check-in camera and a video stream transmission server; Step 4: Obtain baggage check-in images and videos of baggage taken under different viewing areas on the water flow sorting line; Step 5. Deploy the DynQueueViT model for local baggage re-identification feature calculation. The DynQueueViT baggage re-identification model does not store all the features of the training images in the video memory at one time, but uses a first-in-first-out dynamic queue to manage these features. As the training progresses, the features in the queue will be continuously updated to ensure that there are enough samples to participate in the training while significantly reducing the demand for video memory for training. At the same time, during the training process, the DynQueueViT baggage re-identification model uses the time constraint of the video sequence for the training data entering the model, which reduces the noise of pseudo labels to a certain extent. Step 6: Calculate the re-identification features between different luggage using the local luggage re-identification feature calculation model, using the multi-layer visual conversion module and the local feature data enhancement module; Step 7: Substitute the re-identification features between different luggage into the distance formula; determine whether the two target luggages have the same identity.

2. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that: The creation process of the luggage dataset LReID-50tohsouand without identity tags in step 1; the production steps include: using LabelMe and videos to create a small batch detection dataset, initializing and fine-tuning the YOLOv5 network, generating videos marked with luggage bounding boxes, extracting a picture every 4 frames, cropping the bounding boxes to generate a spare dataset, and finally manually screening to exclude duplicate pictures and complete manual labeling of the unlabeled dataset.

3. A completely unsupervised baggage re-identification method based on dynamic queue depth clustering according to claim 1, characterized in that: In step 5, the DynQueueViT model uses a feature extraction network to decompose the luggage image input to the model into global features f g , local features f l Compare feature f with weight reference data .

4. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that: The DynQueueViT model in step 5 is the global feature f of the input image. g and local features f l A first-in-first-out dynamic queue was established respectively; at the same time, in order to capture image information more comprehensively, a dynamic queue for storing mixed features was introduced; when evaluating whether a sample pair belongs to a positive sample pair or a negative sample pair, the DynQueueViT baggage re-identification model mainly relies on global and local feature queues, and when defining the contrast loss to train the feature extraction network, the DynQueueViT baggage re-identification model chooses to use mixed features.

5. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that: In step 5, the DynQueueViT model introduces the time constraint of the video sequence when updating the dynamic queue. Specifically, in an actual baggage sorting line environment, multiple cameras will shoot the baggage at the same time, and the same piece of baggage is likely to be captured by different cameras in the same time period. Therefore, by using the shooting timestamp attached to each image in the LRe-ID50thsouand dataset, the images shot by different cameras in the same time period are classified into the same training batch, and then the training batch images are sent to the DynQueueViT baggage re-identification model for training.

6. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that: In step 5, the DynQueueViT model uses a positive-negative sample matching algorithm to find all luggage images belonging to the same identity in the dynamic queue, and identifies them as a positive sample set; correspondingly, the mixed features obtained by fusing all image features in the positive sample set will be used as the feature representation of the anchor sample; For all images in the dynamic queue {X i ,i=1,2…,M×batch_size}, and calculate their global features f respectively. g and local features f l With all images in the dynamic queue {X j ,j=1,2…,M×batch_size} (including the image itself); the image features in the dynamic queue are assigned to different similarity sets, and the positive and negative sample matching algorithm will also assign the images in the queue to dissimilar sets and boundary sets; those that are similar to image X i The distance is far, located at the total distance S total (X i ,X j ) from the image at the end of ascending sorting, forming image X i The dissimilar set ImageX i The boundary set Then the excluded similar candidate sets Image X j With those in total distance S total (X i ,X j ) in ascending order but not in the candidate set The image composition of In selecting the positive sample set, from image X i Similarity set Select all samples whose distance is less than the threshold value to form a positive sample set In selecting the negative sample set, we take the image X i The boundary set Dissimilar Sets Some pictures in the middle constitute the negative sample set 7. The ViT baggage re-identification method based on local data enhancement according to claim 1, characterized in that: The LocEnViT model in step 5 is constructed by fusing the positive sample set The global and local features of all images in create a comprehensive global feature f i mg and local features f i ml :

8. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that: The DynQueueViT model in step 5 uses a hybrid feature contrast loss L m and data enhancement contrast loss L data Train the model; overall loss function L total It can be expressed as: L total =βL m +(1-β)L data Hybrid feature contrast loss L m The feature extraction network is trained by using a set of positive samples of real data; this loss is further subdivided into two parts: the global mixed feature L mg Contrastive loss and local mixed feature contrast loss L ml : The contrast loss can be expressed as:

9. A completely unsupervised baggage re-identification method based on dynamic queue deep clustering according to claim 1, characterized in that: In step 5, the DynQueueViT model, when determining whether two baggage images belong to the same identity, the LocEnViT baggage re-identification model performs cosine distance calculation:

Images