CN-115240225-B - Pedestrian recognition method, system, device and storage medium based on radio frequency waves

CN115240225BCN 115240225 BCN115240225 BCN 115240225BCN-115240225-B

Abstract

The invention provides a pedestrian recognition method, a system, a device and a storage medium based on radio frequency waves. The method comprises the steps of obtaining a radio frequency thermodynamic diagram group in a target area, simultaneously obtaining pedestrian image data which are captured in the target area according to preset capturing conditions, extracting pedestrian capturing information from the pedestrian image data, mapping the pedestrian capturing information into the radio frequency thermodynamic diagram group to extract a pedestrian radio frequency thermodynamic image group from the radio frequency thermodynamic diagram group, extracting pedestrian features from the pedestrian radio frequency thermodynamic image group, and comparing the extracted pedestrian features with pedestrian features stored in a database to identify whether the captured pedestrian is a warehouse personnel or not. The invention can effectively manage the personnel entering and exiting the warehouse through the method based on the RF-PersonReID, complete statistics of the entering and exiting information and the entering and exiting residence time of the personnel entering and exiting the warehouse, and simultaneously effectively solve the visual interference caused by personal changing and loading or similar dressing.

Inventors

CUI LONG
YOU HAOQUAN
DANG YIFEI
YUAN DESHENG
MA WEIMIN
LIN ZHIQIANG
CHENG XIFENG
WANG HAITAO

Assignees

汇纳科技股份有限公司

Dates

Publication Date: 20260512
Application Date: 20220729

Claims (7)

1. A pedestrian recognition method based on radio frequency waves, comprising: Acquiring first radio frequency thermal image data acquired in a target area through a horizontal antenna array and second radio frequency thermal image data acquired in the target area through a vertical antenna array, and combining the first radio frequency thermal image data and the second radio frequency thermal image data to form a radio frequency thermodynamic diagram group; Meanwhile, pedestrian image data which are captured according to preset capturing conditions in a target area are obtained, and pedestrian capturing information is extracted from the pedestrian image data; A pedestrian detection frame which is snapped is selected by pedestrian detection frames according to the pedestrian image data through a preset pedestrian target detection network, and the pedestrian image data is extracted to contain the snapping time matched with the preset snapping conditions and pedestrian snapping information of the pedestrian detection frame; Mapping the pedestrian snapshot information to a corresponding radio frequency thermodynamic diagram, and digging the pedestrian radio frequency thermodynamic images on the matched radio frequency thermodynamic diagram one by one according to the coordinates of the snapshot pedestrian detection frame to form a pedestrian radio frequency thermodynamic image group; extracting pedestrian characteristics from the pedestrian radio frequency thermal image group; And comparing the extracted pedestrian characteristics with pedestrian characteristics stored in a database to identify whether the snapped pedestrians are warehousing personnel.
2. The method of claim 1, wherein the step of extracting a pedestrian feature vector from the pedestrian rf thermal image comprises: Image processing is carried out on the pedestrian radio frequency thermal image group so as to obtain a pedestrian radio frequency thermal image group with uniform width and height dimensions; and extracting pedestrian characteristics from the pedestrian radio frequency thermal image group after image processing.
3. The radio frequency wave-based pedestrian recognition method according to claim 2, wherein the step of comparing the extracted pedestrian characteristics with pedestrian characteristics stored in the database to recognize whether the snap shot pedestrian is a warehouse-in person comprises: The method comprises the steps of comparing the extracted pedestrian characteristics with pedestrian characteristics stored in a database one by one, identifying the captured pedestrian as a warehousing personnel and marking an entry and exit record and personnel numbers of the captured pedestrian if the comparison is successful, identifying the captured pedestrian as a non-warehousing personnel if the comparison is failed, warehousing the captured pedestrian and assigning personnel numbers to the captured pedestrian.
4. The method for identifying pedestrians based on radio frequency waves according to claim 3, wherein the step of comparing the extracted pedestrian features with the pedestrian features stored in the database one by one in similarity comprises: performing similarity calculation on the pedestrian characteristics and all pedestrian characteristics prestored in a database one by one; when the similarity between the pedestrian characteristic and a pre-stored pedestrian characteristic is greater than or equal to a preset similarity threshold, the comparison is successful; and when the similarity between the pedestrian characteristic and a pre-stored pedestrian characteristic is smaller than a preset similarity threshold value, indicating that the comparison is failed.
5. A radio frequency wave based pedestrian recognition system comprising: the acquisition module is used for acquiring first radio frequency thermal image data acquired in the target area through the horizontal antenna array and second radio frequency thermal image data acquired in the target area through the vertical antenna array, and combining the first radio frequency thermal image data and the second radio frequency thermal image data to form a radio frequency thermodynamic diagram group; The information extraction module is used for detecting pedestrians in the pedestrian image data through a preset pedestrian target detection network, selecting a snapshot pedestrian detection frame in a frame mode, and extracting the snapshot information containing the snapshot time matched with the preset snapshot condition and the pedestrian snapshot information of the pedestrian detection frame from the pedestrian image data; The image matting module is used for mapping the pedestrian snapshot information to a corresponding radio frequency thermodynamic diagram, and one-to-one matting the pedestrian radio frequency thermodynamic image on the matched radio frequency thermodynamic diagram according to the coordinates of the snapshot pedestrian detection frame to form a pedestrian radio frequency thermodynamic image group; the feature extraction module is used for extracting pedestrian features from the pedestrian radio-frequency thermal image group; and the feature comparison module is used for comparing the extracted pedestrian features with the pedestrian features stored in the database so as to identify whether the captured pedestrians are warehousing personnel.
6. A computer-readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the radio frequency wave-based pedestrian recognition method as claimed in any one of claims 1 to 4.
7. The pedestrian recognition device based on the radio frequency waves is characterized by comprising a processor and a memory; the memory is used for storing a computer program; The processor is connected to the memory for executing the computer program stored in the memory to cause the radio frequency wave based pedestrian recognition device to execute the radio frequency wave based pedestrian recognition method of any one of claims 1 to 4.

Description

Pedestrian recognition method, system, device and storage medium based on radio frequency waves Technical Field The invention belongs to the technical field of image processing, relates to a recognition method, and in particular relates to a pedestrian recognition method, a pedestrian recognition system, a pedestrian recognition device and a pedestrian recognition storage medium based on radio frequency waves. Background In modern warehouse management, management in various aspects is involved, wherein warehouse personnel access management is a complex and tedious work, and warehouses are also various due to various kinds of stored articles, such as a refrigeration house, a hazardous chemical substance warehouse, a fresh-keeping warehouse, a constant temperature warehouse and the like. Warehouse operators have strict regulatory requirements and specifications on the number of operator passes and residence times. For each warehouse, a full-time manager is often required to monitor and record the access information of warehouse staff. In the traditional warehouse personnel access management process, RFID (Radio Frequency Identification) technology and manual supervision are mainly adopted. For RFID, a person is required to actively perform device registration or registration, in which case a person uses a non-self certificate or card. For manual supervision, higher labor cost is required, and the possibility of negligence and error making of supervision personnel exists. The prior art has the following technical defects: (1) Before and after the same person goes out of the warehouse, the change of the equipment may happen, and the image-based technology cannot effectively solve the problems of the change of the equipment and the like. (2) Some warehouses require uniform unification or protective clothing to be worn, and individuals who are similarly worn cannot be effectively identified based on the images. (3) The recognition effect is reduced by a drastic change in ambient light. Therefore, how to provide a pedestrian recognition method, system, device and storage medium based on radio frequency waves to solve the defects that the prior art cannot address the visual interference caused by personal changing or similar installation is a technical problem to be solved in the field. Disclosure of Invention In view of the above-mentioned drawbacks of the prior art, an object of the present invention is to provide a method, a system, a device and a storage medium for identifying pedestrians based on radio frequency waves, which are used for solving the problem that the prior art cannot address the visual interference caused by personal changing, changing or similar wearing. In order to achieve the above and other related objects, an aspect of the present invention provides a method for identifying pedestrians based on radio frequency waves, which includes obtaining a radio frequency thermodynamic diagram set in a target area, simultaneously obtaining pedestrian image data captured according to a preset capturing condition in the target area, extracting pedestrian capturing information from the pedestrian image data, mapping the pedestrian capturing information to the radio frequency thermodynamic diagram set to extract a pedestrian radio frequency thermodynamic diagram set from the radio frequency thermodynamic diagram set, extracting pedestrian features from the pedestrian radio frequency thermodynamic diagram set, and comparing the extracted pedestrian features with pedestrian features stored in a database to identify whether the captured pedestrians are warehouse-in personnel. In an embodiment of the invention, the step of acquiring the rf thermodynamic diagram set in the target area includes acquiring first rf thermodynamic image data acquired in the target area through the horizontal antenna array and second rf thermodynamic image data acquired in the target area through the vertical antenna array, and combining the first rf thermodynamic image data and the second rf thermodynamic image data to form the rf thermodynamic diagram set. In one embodiment of the invention, the pedestrian snapshot information comprises snapshot time and a snapshot pedestrian detection frame, and the step of extracting the pedestrian snapshot information from the pedestrian image data comprises the steps of detecting pedestrians in the pedestrian image data through a preset pedestrian target detection network to select the snapshot pedestrian detection frame, and extracting the snapshot time and the snapshot pedestrian detection frame matched with preset snapshot conditions from the pedestrian image data. In one embodiment of the invention, the step of mapping the pedestrian snapshot information to the rf thermodynamic diagram group to extract the pedestrian rf thermodynamic image group from the rf thermodynamic diagram group comprises mapping the snapshot time and the detection frame of the snapshot pedestrian matched with the pre