A multi-source sensing-based tuberculosis sample transfer box monitoring and early warning method and system

Abstract

The present application relates to a kind of based on multi-source sensing's tuberculosis sample transport case monitoring early warning method and system, it is related to the field of transport case safety monitoring, it includes, obtains transport target;Real-time inertia information, real-time temperature and humidity information and continuous video stream data are collected;According to transport target determination box profile parameter;According to real-time inertia information and the preset box inertia correction coefficient determines transient acceleration parameter;Combining continuous video stream data and box profile parameter obtains box surface shape information and internal article state change information;According to transport target, establish safety decision tree, to transient acceleration parameter, box surface shape information, internal article state change information and real-time temperature and humidity information input safety decision tree generates safety grade evaluation result;According to safety grade evaluation result, the corresponding alarm trigger information of output is sent.This application has the effect of effectively monitoring and evaluating the real-time state of article ontology.

Description

Technical Field

[0001] This invention relates to the field of transport box safety monitoring, and in particular to a method and system for monitoring and early warning of tuberculosis sample transport boxes based on multi-source sensing. Background Technology

[0002] Transport container safety monitoring technology is widely used in the transportation of medical supplies such as tuberculosis samples, pathogenic microorganism samples, vaccines, living organs, and stem cells. Among these, sputum samples from tuberculosis patients, Mycobacterium tuberculosis culture samples, and drug sensitivity test samples require frequent transport between sampling points, testing laboratories, and sample banks, placing high demands on biosafety and sample integrity during transport.

[0003] Existing methods for monitoring the safety of transport containers typically rely on single-dimensional scalar sensors such as thermometers, hygrometers, gyroscopes, and accelerometers to collect data on temperature, humidity, attitude angles, and vibration acceleration, and use fixed thresholds for anomaly detection. For example, a vibration alarm is triggered when the vibration value exceeds a preset threshold, or a temperature anomaly alarm is triggered when the temperature deviates from a preset range, thereby indirectly determining whether the transported goods are in an abnormal state.

[0004] However, scalar sensor data typically only reflects changes in the overall environment or orientation of the transport container, making it difficult to directly observe and characterize the real-time state of the items inside. This results in the system being virtually blind in identifying conditions such as item damage, leakage, displacement, tipping over, or contamination. Furthermore, due to the use of a single-threshold judgment mechanism, vibration alarms can be triggered in complex real-world transportation scenarios by prolonged road bumps, speed bumps, or normal loading and unloading impacts, leading to false alarms and difficulty in effectively distinguishing between normal transportation disturbances and genuine safety risks. Consequently, the system cannot effectively monitor and assess the real-time state of the items themselves. Summary of the Invention

[0005] In order to effectively monitor and evaluate the real-time status of the object itself, the present invention provides a monitoring and early warning method and system for tuberculosis sample transport box based on multi-source sensing.

[0006] In a first aspect, the present invention provides a monitoring and early warning method for tuberculosis sample transport boxes based on multi-source sensing, employing the following technical solution:

[0007] A monitoring and early warning method for tuberculosis sample transport boxes based on multi-source sensing, comprising:

[0008] Obtain the transportation target;

[0009] Collect real-time inertial information, real-time temperature and humidity information, and continuous video stream data;

[0010] Determine the container profile parameters based on the transportation objectives;

[0011] The transient acceleration parameters are determined based on real-time inertial information and preset box inertial correction coefficients;

[0012] By combining continuous video stream data and box contour parameters, information on the surface shape of the box and the status changes of the items inside are obtained;

[0013] A safety decision tree is established based on the transportation objectives. Transient acceleration parameters, container surface shape information, internal item status change information, and real-time temperature and humidity information are input into the safety decision tree to generate a safety level assessment result.

[0014] Output the corresponding alarm trigger information based on the security level assessment results.

[0015] By adopting the above technical solution, a corresponding safety decision tree is established based on the transportation target. Real-time inertial information, real-time temperature and humidity information, and continuous video stream data are collected to obtain transient acceleration parameters, container surface shape information, and information on changes in the state of internal items. These data are input into the safety decision tree to generate a safety level assessment result and output corresponding alarm trigger information. This effectively monitors and assesses the real-time state of the items themselves.

[0016] Optionally, obtaining information on the surface shape of the container and changes in the state of the items inside includes:

[0017] Continuous video stream data includes video frames of the outer surface and video frames of the internal object area;

[0018] Perform brightness normalization, filtering and noise reduction, and inter-frame stabilization on the external surface video frames to obtain preprocessed video frames;

[0019] Based on the preset imaging correction parameters, distortion correction is performed on the preprocessed video frames to obtain the corrected video frames;

[0020] Edge detection processing is performed on the corrected video frames to obtain the outer contour edge information of the box.

[0021] The outer contour edge information of the box is matched and compared with the preset box contour parameters to determine the box edge offset, the box surface recessed area and the change of the closed structure.

[0022] Generate the surface shape information of the box based on the offset of the box edge, the recessed area of ​​the box surface, and the change of the closed structure;

[0023] Based on video frames of the internal object area and transient acceleration parameters, internal information is generated and used as information on changes in the state of the internal objects.

[0024] By employing the above technical solution, the outer contour edge information of the container is obtained by processing the surface shape information of the container in continuous video stream data. Based on this outer contour edge information, the surface shape information of the container is determined. This information, combined with video frames of the internal item area and transient acceleration parameters, generates information on the changes in the state of the internal items. This improves the comprehensiveness and reliability of the safety status assessment of the transport container.

[0025] Optionally, the generated internal information includes:

[0026] Transport targets include solid targets and / or liquid targets;

[0027] The impact direction, impact intensity, and impact duration of the transfer box are determined based on transient acceleration parameters.

[0028] When the transport target includes a solid target, target region identification is performed on the video frames of the internal item area to determine the solid detection area;

[0029] Optical flow estimation is performed on continuous solid detection regions to determine the pixel motion direction and pixel displacement of solid targets between adjacent video frames;

[0030] The pixel motion direction and pixel displacement are matched and analyzed with the impact direction, impact intensity and impact duration to generate solid target displacement information;

[0031] When the transport target includes a liquid target, determine the abnormal liquid flow information based on the liquid target;

[0032] Internal information is generated based on at least one of the abnormal liquid flow information and solid target displacement information.

[0033] By employing the above technical solution, solid and liquid targets are distinguished based on the transport target, and the impact direction, intensity, and duration are determined by combining transient acceleration parameters. Through matching analysis of pixel motion direction, pixel displacement, impact direction, impact intensity, and impact duration, abnormal displacement of solid targets and abnormal flow of liquid targets can be identified to generate internal information, thereby improving the accuracy of judging the state of internal items.

[0034] Optionally, abnormal liquid flow information can be determined based on the liquid target, including:

[0035] Determine the visible area of ​​the liquid based on the liquid target;

[0036] Based on the liquid visible area information, target area identification is performed on video frames of the internal object area to determine the liquid detection area;

[0037] Extract the changes in the area of ​​the liquid reflective region, the liquid edge diffusion velocity, and the changes in the liquid flow direction from the liquid detection area;

[0038] By matching and analyzing the changes in the area of ​​the liquid reflective region, the liquid edge diffusion velocity, and the changes in the liquid flow direction with the impact intensity, impact direction, and impact duration, abnormal liquid flow information is generated.

[0039] By adopting the above technical solution, the visible area of ​​the liquid is determined based on the liquid target, and the liquid detection area is identified within this area, thereby improving the targeting of liquid target identification; by extracting the changes in reflective area, edge diffusion velocity and flow direction, and matching and analyzing them with impact intensity, impact direction and impact duration, the accuracy of liquid state monitoring is improved.

[0040] Optionally, generate security level assessment results, including:

[0041] Based on the transportation objectives, the impact threshold, container deformation threshold, internal state change threshold, and temperature and humidity safety range are determined.

[0042] The transient acceleration parameters are compared with the impact threshold to generate an impact state determination result.

[0043] The surface shape information of the enclosure is compared with the deformation threshold of the enclosure to generate the deformation state determination result of the enclosure;

[0044] The internal item status change information is compared with the internal status change threshold to generate an internal item status determination result.

[0045] The real-time temperature and humidity information is compared with the safe temperature and humidity range to generate an environmental condition judgment result.

[0046] The impact condition determination results, enclosure deformation condition determination results, internal item condition determination results, and environmental condition determination results are input into the safety decision tree to generate a result level as the safety level assessment result.

[0047] By adopting the above technical solution, corresponding thresholds and safety ranges can be configured according to the transportation target, and the results of impact, container deformation, internal items and environmental status judgments can be generated respectively. These results are then input into the safety decision tree for comprehensive judgment, thereby achieving multi-dimensional status cross-validation and improving the accuracy of safety level assessment.

[0048] Optionally, the levels of the generated results include:

[0049] When the impact condition determination result is an impact anomaly and the box deformation condition determination result is a box deformation normal, a regular bump condition indicator is generated.

[0050] When the impact condition determination result is impact abnormal and the box deformation condition determination result is box deformation abnormal, a box damage hazard sign is generated.

[0051] When the impact condition determination result is normal and the box deformation condition determination result is abnormal, a crush deformation hazard sign is generated.

[0052] When the internal item status is determined to be abnormal and the environmental status is determined to be normal, an internal movement hazard sign is generated.

[0053] When the internal item status is determined to be abnormal and the environmental status is determined to be abnormal, a hazard sign for internal item damage and environmental out-of-control is generated.

[0054] The result level is generated based on at least one of the following: conventional bump condition warning, container damage hazard warning, crushing deformation hazard warning, internal movement hazard warning, internal item damage and environmental out-of-control hazard warning.

[0055] By adopting the above technical solution, the impact state, container deformation state, internal item state and environmental state are input into the safety decision tree for combined judgment. This can distinguish different risk situations such as normal bumps, container damage, extrusion deformation, internal movement, internal item damage and environmental out-of-control, and generate corresponding result levels. This avoids misjudgment caused by single threshold alarms and improves the accuracy of transport container safety assessment and hierarchical alarm capabilities.

[0056] Optionally, generating internal movement hazard signs includes:

[0057] When the internal item status assessment result is that the solid target displacement is abnormal, and the environmental status assessment result is that the environment is normal, a solid displacement hazard sign is generated.

[0058] When the internal item status assessment result indicates abnormal liquid flow and the environmental status assessment result indicates normal environment, a hidden leak hazard sign is generated.

[0059] An internal movement hazard sign is generated when at least one of a solid displacement hazard sign and a hidden leakage hazard sign is present.

[0060] By adopting the above technical solution, it is possible to distinguish between abnormal displacement of solid targets and abnormal flow of liquid targets. This allows for further differentiation between abnormal displacement of solid targets and hidden leakage of liquid targets even when the environment is not yet abnormal, thereby improving the early identification capability of internal anomalies.

[0061] Optionally, the levels of the generated results include:

[0062] When only a regular bump condition indicator is generated, the safety level assessment result is determined to be low risk.

[0063] When an extrusion deformation hazard sign is generated, the safety level assessment result is determined to be of medium risk level;

[0064] When a hazard sign for damage to the enclosure or an hazard sign for internal movement is generated, the safety level assessment result is determined to be a high-risk level.

[0065] When a hazard sign indicating damage to internal items and loss of environmental control is generated, the safety assessment result is determined to be a severe risk level.

[0066] The corresponding result level is generated based on the low risk level, medium risk level, high risk level, or severe risk level.

[0067] By adopting the above technical solution, different status identifiers can be mapped to low-risk, medium-risk, high-risk, and severe-risk levels, and alarm trigger information of corresponding intensity can be output according to the risk level, thereby realizing graded response, avoiding the triggering of high-intensity alarms by slight bumps, and improving the accuracy and pertinence of alarm processing.

[0068] Optionally, the output alarm trigger information corresponding to the intensity includes:

[0069] When the safety level assessment result is low risk, a status recording signal is generated, and the corresponding real-time inertial information, real-time temperature and humidity information, and continuous video stream data are cached and recorded.

[0070] When the safety level assessment result is medium risk level, a level 1 alarm signal is generated, and an abnormal prompt message is output based on the level 1 alarm signal, which is also used as the alarm trigger information.

[0071] When the safety level assessment result is high risk level, a level 2 alarm signal is generated, and the location of the abnormal area of ​​the cabinet or the abnormal area of ​​the internal items is determined based on continuous video stream data, and the location of the abnormal area of ​​the cabinet or the abnormal area of ​​the internal items is output according to the level 2 alarm signal as alarm trigger information.

[0072] When the safety level assessment result is a severe risk level, a level 3 alarm signal is generated and an alarm is executed based on the level 3 alarm signal. At the same time, images of abnormal areas with corresponding internal item damage or liquid leakage are obtained based on continuous video stream data, and abnormal area images, real-time inertial information, and real-time temperature and humidity information are uploaded as alarm trigger information.

[0073] By adopting the above technical solution, alarm trigger information of different intensities can be output according to the security level; when the risk is low, only data is recorded; when the risk is medium, an anomaly prompt is output; when the risk is high, the abnormal area is located and output; when the risk is severe, an alarm is executed and abnormal images and sensor data are uploaded, thereby realizing risk-level response and abnormal evidence retention, improving alarm handling efficiency and traceability.

[0074] Secondly, the present invention provides a tuberculosis sample transport box monitoring and early warning system based on multi-source sensing, which adopts the following technical solution:

[0075] A monitoring and early warning system for tuberculosis sample transport boxes based on multi-source sensing, comprising:

[0076] The acquisition module is used to acquire real-time inertial information, real-time temperature and humidity information, and continuous video stream data;

[0077] The memory stores a program for implementing a control method for real-time monitoring and evaluation of the safety of a transport container based on multi-source sensing, as described in any one of the above descriptions.

[0078] The processor loads and executes programs stored in memory.

[0079] By employing the above technical solution, real-time inertial information, real-time temperature and humidity information, and continuous video stream data are collected by the acquisition module. The processor then executes the program stored in the memory to obtain transient acceleration parameters, surface shape information of the enclosure, and information on changes in the state of the internal items. This information is input into a safety decision tree to generate a safety level assessment result and outputs corresponding alarm trigger information. This effectively monitors and assesses the real-time state of the items themselves.

[0080] In summary, the present invention has at least one of the following beneficial technical effects:

[0081] 1. Establish a safety decision tree based on transportation objectives, and collect real-time inertial information, real-time temperature and humidity information and continuous video stream data to obtain transient acceleration parameters, container surface shape information and internal item status change information, thereby effectively monitoring and evaluating the real-time status of the item itself;

[0082] 2. By identifying edge offsets, depressions, and changes in the closed structure through video recognition on the outer surface of the enclosure, and combining this with internal video recognition of the item's status, the ability to identify anomalies in the enclosure and the item itself is improved.

[0083] 3. Output alarm trigger information of corresponding intensity according to different risk levels, so that normal bumps are only recorded, while high-risk bumps are promptly alarmed and abnormal data is uploaded, thereby improving the efficiency of handling and tracing. Attached Figure Description

[0084] Figure 1 This is a flowchart of a tuberculosis sample transport box monitoring and early warning method based on multi-source sensing;

[0085] Figure 2 It is a flowchart for obtaining information on the surface shape of the box and the state changes of the items inside. Detailed Implementation

[0086] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0087] This invention discloses a monitoring and early warning method for tuberculosis sample transport boxes based on multi-source sensing. The method involves acquiring the items to be transported as the transport target and establishing a safety decision tree based on the transport target. Real-time inertial information, real-time temperature and humidity information, and continuous video stream data are acquired. The size and shape information of the transport box are determined based on the transport target as the box's contour parameters. The total weight of the transport box carrying the transport target is used as the box's inertial correction coefficient, and transient acceleration parameters are determined in conjunction with the real-time inertial information. The surface shape information of the transport box and the internal state change information of the items inside the transport box are obtained through continuous video stream data and the box's contour parameters. The determined transient acceleration parameters, box surface shape information, internal item state change information, and real-time temperature and humidity information are input into the safety decision tree to generate a safety level assessment result. Based on the safety level assessment result, corresponding alarm trigger information is output, thereby effectively monitoring and assessing the real-time state of the items themselves.

[0088] Reference Figure 1 A monitoring and early warning method for tuberculosis sample transport boxes based on multi-source sensing includes:

[0089] S100: Obtain the transportation target.

[0090] Transport targets refer to the types of goods, packaging forms, and transport safety requirements of goods loaded in transshipment containers and requiring safety monitoring during transport. Transport targets include solid targets and / or liquid targets.

[0091] Obtain the item type, packaging form, and transportation safety requirements of the items to be transferred, and designate the items to be transferred as the transportation target.

[0092] In this embodiment, the transport target can be at least one of the following: sputum samples from tuberculosis patients, culture samples of Mycobacterium tuberculosis, samples for drug sensitivity testing, and samples preserved in a biobank. For tuberculosis sample transport scenarios, the transport target typically adopts a three-layer packaging structure that meets the UN3373 Class B transport requirements for biological materials, including a main container, auxiliary packaging, and outer packaging. The main container is used to store the sample, the auxiliary packaging is used to prevent leakage and absorb any leaked liquid, and the outer packaging is used to provide mechanical protection.

[0093] S101: Collects real-time inertial information, real-time temperature and humidity information, and continuous video stream data.

[0094] Real-time inertial information refers to the acceleration, angular velocity, and attitude changes of the transport container as it moves with the transport vehicle during transportation. Real-time temperature and humidity information refers to the real-time temperature and humidity information of the storage area inside the transport container and / or the transport vehicle compartment. Continuous video stream data refers to continuous image data used to characterize the state of the transport container's outer surface and the state of the contents inside.

[0095] The MPU-6050 inertial sensor, installed inside the transport container, its mounting base, or the transport compartment and kept relatively fixed to the transport container, collects the acceleration, angular velocity, and attitude changes of the transport container during transportation to obtain real-time inertial information. The SHT31 temperature and humidity sensor, installed in the storage area inside the transport container and in the transport compartment, collects temperature and humidity data to obtain real-time temperature and humidity information. The OV5640 external camera, installed inside the transport compartment and facing the outer surface of the transport container, collects video streams of the outer surface of the transport container, and the OV5640 internal camera, installed inside the transport container or facing the transparent observation window of the container, collects video streams of the area containing the contents inside the container. The video streams of the outer surface of the transport container and the area containing the contents inside the container are used as continuous video stream data.

[0096] In this embodiment, after collecting real-time inertial information, real-time temperature and humidity information, and continuous video stream data, time synchronization processing is performed on the real-time inertial information, real-time temperature and humidity information, and continuous video stream data. Time synchronization processing refers to unifying the data collected by inertial sensors, temperature and humidity sensors, external cameras, and internal cameras to the same time reference, so that inertial changes, visual changes, and environmental changes corresponding to the same transportation event can be correlated and analyzed.

[0097] In this embodiment, coordinate mapping processing is also performed on the continuous video stream data and real-time inertial information. Coordinate mapping processing refers to converting the pixel motion direction in the video frame of the internal object area into the relative motion direction in the box coordinate system according to the preset calibration relationship between the internal camera coordinate system, the box coordinate system and the inertial sensor coordinate system, and converting the acceleration vector direction collected by the inertial sensor into the impact direction in the box coordinate system.

[0098] The preset calibration relationships include the installation angle of the internal camera relative to the enclosure, the mapping relationship between the internal camera's imaging plane and the enclosure's reference plane, and the mapping relationship between the inertial sensor's installation direction and the enclosure's reference coordinate system. By using these preset calibration relationships, the pixel movement direction and impact direction can be compared under the enclosure's coordinate system, thereby reducing orientation judgment errors caused by the camera's movement with the enclosure.

[0099] S102: Determine the container profile parameters based on the transportation objectives.

[0100] Box profile parameters refer to the reference parameters used to characterize the external profile features of the transport box in a normal closed state, including box specifications and dimensions, outer profile edge coordinates, box corner positions, lid gap reference width, latch reference position, and closure structure reference position.

[0101] The specifications of the transshipment container used to load the transported goods are determined based on the type of goods, packaging form, and transport safety requirements of the transported goods, and the container outline parameters corresponding to the transshipment container specifications are called.

[0102] In this embodiment, when the transport target is a tuberculosis sample, the box contour parameters also include the preset installation position and fixed structure position of the sample packaging components, which are used to monitor the displacement of the sample container during transportation.

[0103] S103: Determine the transient acceleration parameters based on real-time inertial information and preset box inertial correction coefficients.

[0104] The preset container inertia correction coefficient is determined based on the container specifications, target weight, container installation method, sensor installation location, and container center of gravity location. Transient acceleration parameters refer to the short-term dynamic acceleration parameters of the container during transportation caused by sudden stops, sharp turns, collisions, loading / unloading impacts, or road bumps.

[0105] The triaxial acceleration values ​​are obtained based on real-time inertial information, and the transient acceleration parameters are calculated using the transient acceleration parameter calculation formula, combined with the box inertia correction coefficient. The transient acceleration parameter calculation formula is as follows:

[0106] A t =K b *|(a 2 x +a 2 y +a 2 z ) 1 / 2 -g|;

[0107] A t K is a transient acceleration parameter. b a is the box inertia correction factor. x a y a z The values ​​are the triaxial accelerations collected by the inertial sensor, where g is the gravitational acceleration constant, which is taken as 9.8 m / s² in this embodiment. 2 .

[0108] S104: Combine continuous video stream data and box contour parameters to obtain box surface shape information and internal item status change information.

[0109] Box surface shape information refers to information used to characterize changes in the structural state of the outer surface of the transport box, including the offset of the box edge, the recessed area of ​​the box surface, the change in the gap of the box lid, the change in the position of the latch, and the change in the continuity of the closed structure.

[0110] Internal item state change information refers to information used to characterize changes in the position, attitude, flow, or leakage of transported targets inside the transshipment container during transportation. This includes solid target displacement information and abnormal liquid flow information. Solid target displacement information includes the pixel movement direction, pixel displacement amount, centroid offset, attitude change amount, and trajectory change amount. Abnormal liquid flow information includes changes in the area of ​​the liquid reflective region, liquid edge diffusion velocity, and changes in liquid flow direction.

[0111] External video frames of the transport container are extracted from the continuous video stream data. Edge matching and contour comparison are performed between the external video frames and the container's contour parameters to obtain the surface shape information of the container. Internal video frames are extracted from the continuous video stream data, and visual-inertial matching analysis is performed on the internal video frames in conjunction with transient acceleration parameters to obtain information on changes in the state of the internal items.

[0112] S105: Establish a safety decision tree based on the transportation objective, and input transient acceleration parameters, container surface shape information, internal item status change information and real-time temperature and humidity information into the safety decision tree to generate a safety level assessment result.

[0113] A safety decision tree is a pre-configured multi-node branching logic based on the type of goods being transported, the packaging format, and the transportation safety requirements. It is used to cross-reference transient acceleration parameters, container surface shape information, internal item status changes, and real-time temperature and humidity information. The safety level assessment result refers to the risk level output by the safety decision tree based on the cross-reference results of multi-source sensor data.

[0114] The required safety thresholds are determined based on the transportation objectives, and a safety decision tree is built based on these thresholds. The safety decision tree includes nodes for impact state assessment, container deformation state assessment, internal item state assessment, and environmental state assessment. Transient acceleration parameters, container surface shape information, internal item state change information, and real-time temperature and humidity information are input into the safety decision tree. The safety decision tree then generates the respective state assessment result nodes, which are combined to produce a safety level assessment result.

[0115] S106: Output the corresponding alarm trigger information based on the security level assessment results.

[0116] Alarm triggering information refers to information used to trigger alarm actions and characterize the cause of the anomaly.

[0117] Based on the security level assessment results, the corresponding alarm intensity and alarm method are determined, and corresponding alarm trigger information is generated to effectively monitor and assess the real-time status of the item itself.

[0118] Reference Figure 2 The information obtained includes the shape of the box surface and the status changes of the items inside.

[0119] S200: Continuous video stream data includes video frames of the outer surface and video frames of the internal object area.

[0120] External surface video frames refer to image frames continuously acquired from the external surface of the transport container, used to characterize the image status of the external outline, container surface, lid gaps, locking structure, and closure structure of the transport container. Internal item area video frames refer to image frames continuously acquired from the internal item storage area of ​​the transport container, used to characterize the placement status, pose changes, liquid flow, or leakage status of the transported items inside the transport container.

[0121] S201: Perform brightness normalization, filtering and noise reduction, and inter-frame stabilization processing on the external surface video frames to obtain preprocessed video frames.

[0122] Brightness normalization refers to adjusting the overall brightness or grayscale distribution of video frames on the outer surface to ensure that video frames acquired at different times are within a relatively consistent brightness range. Filtering and denoising refers to removing random noise, granular noise, and small-area interference points from video frames on the outer surface using at least one of Gaussian filtering, median filtering, or bilateral filtering. Inter-frame stabilization refers to methods that perform translation, rotation, and scale compensation on video frames based on feature point matching results or image motion estimation results between adjacent video frames.

[0123] Brightness normalization, filtering and noise reduction, and inter-frame stabilization are sequentially performed on the external surface video frames to obtain pre-processed video frames, thereby reducing the impact of transport vehicle vibration or camera shake on the external contour recognition of the box.

[0124] S202: Perform distortion correction on the preprocessed video frame according to the preset imaging correction parameters to obtain the corrected video frame.

[0125] Preset imaging correction parameters refer to calibration parameters used to correct camera imaging distortion, including at least one of camera intrinsic parameters, radial distortion parameters, and tangential distortion parameters. Distortion correction refers to the image processing procedure that corrects image curvature, scaling distortion, or positional shift in pre-processed video frames caused by lens distortion, mounting deviation, or imaging plane tilt. A corrected video frame is the video frame obtained after distortion correction.

[0126] Distortion correction is performed on the preprocessed video frames according to preset imaging correction parameters to obtain corrected video frames. This ensures that the edge positions and outline proportions of the container in the corrected video frames closely approximate the actual appearance of the transport container.

[0127] S203: Perform edge detection processing on the corrected video frame to obtain the outer contour edge information of the box.

[0128] Edge detection processing refers to the image processing procedure of identifying regions with grayscale, color, or texture changes in a corrected video frame to extract the outer contour boundary of the transport container. This includes at least one of Canny edge detection, Sobel edge detection, Laplacian edge detection, and Hough line detection. The outer contour edge information of the container refers to the data used to characterize the outer contour boundary of the transport container, including container edge segments, container corner positions, container contour coordinates, and container edge continuity.

[0129] Edge detection processing is performed on the corrected video frame to extract the edge segments and corner positions corresponding to the outer contour of the transport box, and the edge segments and corner positions are used as the outer contour edge information of the box.

[0130] S204: Match and compare the outer contour edge information of the box with the preset box contour parameters to determine the box edge offset, the box surface recessed area and the change of the closed structure.

[0131] The preset box profile parameters refer to the external profile reference parameters of the transport box under normal closed and undamaged conditions, including the reference edge coordinates of the box, the reference corner position, the reference width of the box lid gap, the reference position of the latch, and the reference continuity parameters of the seal.

[0132] The offset of the enclosure edge refers to the offset of the real-time edge coordinates or real-time corner position in the outer contour edge information of the enclosure relative to the reference edge coordinates or reference corner position in the enclosure contour parameters. The recessed area on the enclosure surface refers to an abnormally recessed area on the enclosure surface. The change in the closure structure refers to the change in the gaps in the enclosure lid, latches, hinges, seals, or other closure structures relative to the corresponding reference position or reference continuity parameter.

[0133] The real-time obtained outer contour edge information of the box is matched and compared with the box contour parameters. The offset of the box edge is determined according to the offset of the real-time edge coordinates or real-time corner position relative to the corresponding reference parameter. The abnormal area of ​​the box surface is determined according to the local contour deviation, edge curvature change or surface texture change between the outer contour edge information of the box and the box contour parameters. The change in the sealing structure is determined according to the change of the box lid gap, the position of the latch or the continuity of the seal.

[0134] S205: Generate the surface shape information of the box based on the offset of the box edge, the recessed area of ​​the box surface, and the change of the closed structure.

[0135] Box surface shape information refers to comprehensive feature information used to characterize the structural state of the outer surface of the transport box, including the offset of the box edge, the recessed area of ​​the box surface, the amount of change in the closed structure and its corresponding location and degree of change.

[0136] The offset of the box edge, the concave area of ​​the box surface, and the change in the closed structure are combined to generate the shape information of the box surface. The combined processing includes at least one of feature normalization, feature encoding, or feature vector generation.

[0137] S206: Generate internal information based on video frames of the internal object area and transient acceleration parameters, and use it as information on changes in the state of the internal objects.

[0138] Internal information refers to comprehensive information used to characterize changes in the position, orientation, movement, flow, or leakage of transported objects inside the transshipment container during transportation.

[0139] Target region identification and inter-frame change analysis are performed on video frames of the internal object area to obtain the visual state change characteristics of the internal objects. The visual state change characteristics are then matched and analyzed with transient acceleration parameters to generate internal information, which is used as the internal object state change information.

[0140] The generated internal information includes:

[0141] S300: Transport targets include solid targets and / or liquid targets.

[0142] Solid targets refer to transport targets that have a relatively fixed shape during transportation, including instruments, equipment, packaging boxes, reagent bottles, sample tubes, and organ preservation containers. Liquid targets refer to transport targets that are fluid or subject to liquid level changes during transportation, including preservation solutions, buffer solutions, culture media, refrigerants, biochemical reagents, or leaked liquids.

[0143] The transport target includes at least one of a solid target and / or a liquid target.

[0144] S301: Determine the impact direction, impact intensity, and impact duration of the transfer box based on the transient acceleration parameters.

[0145] The impact direction refers to the direction of the acceleration vector corresponding to the transient impact on the transport container, and is used to characterize the main direction of the impact on the transport container. The impact intensity refers to the peak value of the transient acceleration parameter during the impact process, and is used to characterize the strength of the impact. The impact duration refers to the length of time that the transient acceleration parameter continuously exceeds a preset impact threshold, and is used to characterize the duration of the impact.

[0146] The transient acceleration vector is determined based on the triaxial acceleration components in real-time inertial information, and the impact direction is determined based on the direction of the transient acceleration vector. The impact intensity is determined based on the peak value of the transient acceleration parameter. The impact duration is determined based on the time difference between the start and end times when the transient acceleration parameter continuously exceeds a preset impact threshold. The preset impact threshold is the acceleration threshold used to determine whether the transfer box has entered an impact state.

[0147] S302A: When the transport target includes a solid target, target area identification is performed on the video frames of the internal item area to determine the solid detection area.

[0148] The solid detection region refers to the image region in the video frame corresponding to the solid target, which is used to detect the position, contour and motion changes of the solid target.

[0149] When the transport target includes a solid target, target region recognition processing is performed on the video frames of the internal item area to identify the target contour and target bounding box corresponding to the solid target, and the target contour and target bounding box are determined as the solid detection area.

[0150] S302A1: Performs optical flow estimation processing on continuous solid detection areas to determine the pixel motion direction and pixel displacement of solid targets between adjacent video frames.

[0151] Optical flow estimation refers to the process of estimating the motion vector of a solid target in the image coordinate system based on the changes in pixel grayscale and feature points within the solid detection region in adjacent video frames. The pixel motion direction refers to the orientation of the solid target in the image coordinate system. The pixel displacement refers to the pixel-level displacement of the solid target along the pixel motion direction between adjacent video frames.

[0152] Optical flow estimation is performed on continuous solid detection areas to obtain the motion vector of the solid target between adjacent video frames, and the pixel motion direction and pixel displacement of the solid target are determined based on the motion vector.

[0153] S302A2: Matches and analyzes the pixel motion direction and pixel displacement with the impact direction, impact intensity and impact duration to generate solid target displacement information.

[0154] Solid target displacement information refers to information used to characterize the positional changes, attitude changes, or abnormal displacement states of solid targets inside the transport container during transportation.

[0155] Matching analysis refers to the analysis of the visual state change characteristics of the transport target in the video frame of the internal goods area and the transient impact change of the transfer box in terms of time correspondence, direction correspondence and amplitude correspondence.

[0156] When the pixel motion direction and the impact direction satisfy a preset direction correspondence, and the pixel displacement is within the allowable displacement range corresponding to the impact intensity, at least one of the pixel motion direction, pixel displacement, impact direction, impact intensity, solid detection area position, and corresponding timestamp is encapsulated as inertial sway displacement information; when the pixel motion direction and the impact direction do not satisfy the preset direction correspondence, or the pixel displacement exceeds the allowable displacement range, or the pixel does not recover to the initial position range or stable state within a preset recovery time after the impact duration ends, at least one of the pixel motion direction, pixel displacement, impact direction, impact intensity, abnormal duration, solid detection area position, and corresponding timestamp is encapsulated as abnormal displacement information; the inertial sway displacement information and / or abnormal displacement information are used as solid target displacement information.

[0157] The preset orientation correspondence refers to the orientation matching rules predetermined based on the coordinate mapping relationship between the internal camera coordinate system, the box coordinate system, and the inertial sensor coordinate system, combined with the inertial correspondence between the box's motion direction and the relative motion direction of the internal solid target. The preset orientation correspondence is used to determine whether the relative motion direction of the solid target in the box coordinate system corresponds to the impact direction of the transport box.

[0158] S302B: When the transport target includes a liquid target, determine abnormal liquid flow information based on the liquid target.

[0159] Information on abnormal liquid flow refers to information used to characterize abnormal shaking, diffusion, leakage, or changes in the flow state of liquid targets inside the transport container during transportation. This includes changes in the area of ​​the liquid reflective zone, the diffusion rate of the liquid edge, changes in the liquid flow direction, and the abnormal liquid flow state.

[0160] When the transport target includes a liquid target, the abnormal liquid flow information corresponding to the liquid target is determined based on the packaging form of the liquid target and the liquid visibility area information.

[0161] S302: Generate internal information based on at least one of the abnormal liquid flow information and the solid target displacement information.

[0162] When solid target displacement information is available, it indicates changes in the position, attitude, or trajectory of the solid target inside the transfer container; when abnormal liquid flow information is available, it indicates changes in the flow or leakage status of the liquid target inside the transfer container; solid target displacement information and / or abnormal liquid flow information are used as internal information.

[0163] Determine abnormal liquid flow information based on the liquid target, including:

[0164] S400: Determines liquid visibility area information based on liquid targets.

[0165] Liquid visibility area information refers to the image area information in the video frame of the internal object area that can be used to observe changes in the state of the liquid target, including transparent container area, semi-transparent container area, liquid level observation window area, liquid leakage receiving area and liquid absorption pad area.

[0166] When the transport target includes a liquid target, the visible area corresponding to the liquid target in the video frame of the internal item area is determined based on the packaging form and visible structure information of the liquid target, and the visible area is used as the liquid visible area information.

[0167] S401: Based on the liquid visible area information, target area recognition is performed on the video frames of the internal object area to determine the liquid detection area.

[0168] The liquid detection area refers to the image area in the video frame corresponding to the liquid target, which is used to detect the position, edge, reflective features and motion changes of the liquid target.

[0169] Based on the visible liquid area information, the candidate recognition range of the liquid target is determined in the video frame of the internal object area, and the target area recognition processing is performed within the candidate recognition range to identify the target contour and target bounding box area corresponding to the liquid target. The liquid surface boundary position of the liquid target is determined according to the target contour, and the liquid surface boundary position and the target bounding box area are jointly determined as the liquid detection area.

[0170] S402: Extract the change in the area of ​​the liquid reflective region, the diffusion velocity at the liquid edge, and the change in the direction of liquid flow from the liquid detection area.

[0171] The change in the area of ​​the liquid reflective region refers to the change in the area of ​​the liquid reflective region over time in consecutive video frames. The liquid edge diffusion velocity refers to the speed at which the liquid boundary expands or moves outward in consecutive video frames. The change in liquid flow direction refers to the change in the direction of liquid flow, edge expansion, or reflective region movement in consecutive video frames.

[0172] Brightness features are extracted and regions are segmented in the liquid detection area to determine the liquid reflective area. Based on the changes in the reflective area, liquid boundary position and liquid movement direction in adjacent video frames, the changes in the reflective area, liquid edge diffusion velocity and liquid flow direction are determined respectively.

[0173] In this embodiment, the liquid detection area is subjected to HSV color space conversion, brightness threshold segmentation, background difference and / or target mask segmentation to determine the liquid reflection area; the change in the area of ​​the liquid reflection area is determined based on the difference or rate of change of the area of ​​the liquid reflection area in consecutive video frames.

[0174] S403: Match the changes in the area of ​​the liquid reflective region, the liquid edge diffusion velocity, and the changes in the liquid flow direction with the impact intensity, impact direction, and impact duration to generate abnormal liquid flow information.

[0175] Information on abnormal liquid flow refers to information used to characterize abnormal shaking, continuous diffusion, leakage, or abnormal changes in the flow direction of liquid targets inside the transport container during transportation.

[0176] The changes in the area of ​​the liquid reflective region, the diffusion velocity of the liquid edge, and the change in the liquid flow direction are matched and analyzed with the impact intensity, impact direction, and impact duration. When the change in the liquid flow direction and the impact direction meet the preset direction correspondence, and the changes in the area of ​​the liquid reflective region and the diffusion velocity of the liquid edge are within the allowable range corresponding to the impact intensity, and recover to the preset stable range within the preset recovery time after the end of the impact duration, the process jumps to collect real-time inertial information, real-time temperature and humidity information, and continuous video stream data.

[0177] When the change in liquid flow direction does not satisfy the preset correspondence between the impact direction and the liquid flow direction, or when the change in the area of ​​the liquid reflective region continues to increase, or when the liquid edge diffusion speed exceeds the allowable range, or when the liquid does not recover to the preset stable range within the preset recovery time, the above judgment result is generated. The result is then combined with the information on the change in liquid flow direction and impact direction, the change in the area of ​​the liquid reflective region and the recovery time to form the abnormal liquid flow information.

[0178] Generate security level assessment results, including:

[0179] S500: Determines the impact threshold, container deformation threshold, internal state change threshold, and temperature and humidity safety range based on the transportation target.

[0180] Impact threshold refers to the threshold used to determine whether a transport container has been subjected to abnormal impacts during transportation. Container deformation threshold refers to the threshold used to determine whether changes in the container's edge offset, surface dents, or enclosed structure are abnormal. Internal state change threshold refers to the threshold used to determine whether changes in the state of the internal items are abnormal. Temperature and humidity safe range refers to the threshold used to determine whether real-time temperature and humidity information is within a safe range.

[0181] An acceleration threshold predetermined based on the type of goods, packaging form, and impact resistance requirements of the transported item is used as the impact threshold. A pre-determined allowable deformation range of the outer surface of the container, based on the specifications of the transported container and transport safety requirements, is used as the container deformation threshold. Pre-determined solid target displacement thresholds and liquid flow anomaly thresholds, based on the fixing method, packaging form, and item condition requirements of the transported item, are used as internal state change thresholds. Pre-determined allowable temperature and humidity ranges, based on the storage conditions and transport safety requirements of the transported item, are used as the temperature and humidity safety range.

[0182] In this embodiment, when the transport target is a tuberculosis sample, the safe temperature and humidity range is preset according to the sample preservation requirements. For example, the temperature is controlled within the range of 2°C to 8°C during the transport of sputum samples to reduce the risk of sample inactivation and contamination.

[0183] S501: Compare the transient acceleration parameter with the impact threshold to generate the impact state determination result.

[0184] The impact condition determination result refers to the determination result used to characterize whether the transfer container has been subjected to abnormal impact during transportation, including normal impact and abnormal impact.

[0185] The transient acceleration parameter is compared with the impact threshold. When the transient acceleration parameter does not exceed the impact threshold, an impact state judgment result is generated to characterize the normal impact. When the transient acceleration parameter exceeds the impact threshold, an impact state judgment result is generated to characterize the abnormal impact.

[0186] S502: Compare the surface shape information of the enclosure with the enclosure deformation threshold to generate the enclosure deformation state determination result.

[0187] The container deformation status determination result refers to the determination result used to characterize whether the container has undergone abnormal deformation during transportation, including normal container deformation and abnormal container deformation.

[0188] The box edge offset, box surface depression area and closed structure change in the box surface shape information are compared with the corresponding box deformation threshold. When the box surface shape information does not exceed the corresponding box deformation threshold, a box deformation state judgment result is generated to characterize the box deformation as normal. When the box surface shape information exceeds the corresponding box deformation threshold, a box deformation state judgment result is generated to characterize the box deformation as abnormal.

[0189] S503: Compare the internal item status change information with the internal status change threshold to generate an internal item status determination result.

[0190] The internal item status determination result refers to the determination result used to characterize whether the internal items of the transfer box have undergone abnormal status changes during transportation, including internal items being normal and internal items being abnormal.

[0191] The solid target displacement information and / or abnormal liquid flow information in the internal item state change information are compared with the corresponding internal state change thresholds. When the internal item state change information does not exceed the corresponding internal state change threshold, an internal item state determination result is generated to characterize the internal item as normal. When the internal item state change information exceeds the corresponding internal state change threshold, an internal item state determination result is generated to characterize the internal item as abnormal.

[0192] S504: Compares real-time temperature and humidity information with the safe temperature and humidity range to generate an environmental condition judgment result.

[0193] The environmental condition determination result refers to the determination result used to characterize whether the internal temperature and humidity of the transshipment container meet the preservation requirements of the transport target, including normal environment and abnormal environment.

[0194] The real-time temperature and humidity values ​​in the real-time temperature and humidity information are compared with the temperature safety range and humidity safety range, respectively. When the real-time temperature value is within the temperature safety range and the real-time humidity value is within the humidity safety range, an environmental state judgment result is generated to characterize the normal environment. When the real-time temperature value exceeds the temperature safety range and / or the real-time humidity value exceeds the humidity safety range, an environmental state judgment result is generated to characterize the abnormal environment.

[0195] S505: Input the impact state determination results, enclosure deformation state determination results, internal item state determination results, and environmental state determination results into the safety decision tree to generate a result level as the safety level assessment result.

[0196] The result rating refers to the rating used to characterize the current level of safety risk of the transport container.

[0197] The impact state determination results, the enclosure deformation state determination results, the internal item state determination results, and the environmental state determination results are input into the safety decision tree. The safety decision tree generates a result level based on the combination relationship between different determination results, which serves as the safety level assessment result.

[0198] The generated result levels include:

[0199] S600: When the impact condition determination result is an impact anomaly and the box deformation condition determination result is a box deformation normal, a normal bump condition indicator is generated.

[0200] The standard bump condition indicator refers to the status indicator when the transshipment container is subjected to a short-term impact or road bumps during transportation, but no abnormal deformation occurs on the outer surface of the container.

[0201] When the impact condition determination result is an abnormal impact and the box deformation condition determination result is a normal box deformation, it indicates a short-term impact on the transfer box during transportation. However, the offset of the box edge, the indentation area on the box surface, or the change in the closed structure does not reach the abnormal deformation condition, so a normal bumpy condition indicator is generated.

[0202] In this embodiment, a normal transportation status identifier is generated when the impact status determination result is normal, the box deformation status determination result is normal, the internal item status determination result is normal, and the environmental status determination result is normal. The normal transportation status identifier refers to the status identifier generated when no abnormal impact, box deformation, internal item abnormality, or temperature / humidity environmental abnormality is detected during the transport of the transshipment box.

[0203] S601: When the impact condition determination result is impact abnormal and the box deformation condition determination result is box deformation abnormal, a box damage hazard sign is generated.

[0204] The container damage hazard label indicates a danger when the transshipment container is subjected to abnormal impact during transportation, resulting in abnormal deformation of the outer surface of the container.

[0205] When the impact condition determination result is an impact anomaly, and the box deformation condition determination result is a box deformation anomaly, it indicates that the transport box has been subjected to an abnormal impact, and the outer surface of the box shows edge displacement, surface dent, or abnormal changes in the closed structure, thus generating a box damage hazard sign.

[0206] S602: When the impact condition determination result is normal and the box deformation condition determination result is abnormal, a crush deformation hazard sign is generated.

[0207] The extrusion deformation hazard sign indicates a danger sign when abnormal deformation occurs on the outer surface of the transport container without the detection of obvious transient impact.

[0208] When the impact condition is determined to be normal and the deformation condition is determined to be abnormal, it indicates that the transport container may be subjected to continuous compression, stacking, prying or slow external force, thus generating a compression deformation hazard sign.

[0209] S603: When the internal item status determination result is that the internal item is abnormal, and the environmental status determination result is that the environment is normal, an internal movement hazard sign is generated.

[0210] Internal movement hazard signs indicate that there is abnormal displacement, abnormal flow, or suspected leakage of items inside the transport container, but the temperature and humidity environment inside the transport container has not exceeded the safe range.

[0211] When the internal item status is determined to be abnormal and the environmental status is determined to be normal, it indicates that the internal items of the transport container have an abnormal position, posture, flow or leakage status, but the abnormality has not yet caused the temperature and humidity environment to run out of control, so an internal movement hazard sign is generated.

[0212] S604: When the internal item status determination result is that the internal item is abnormal, and the environmental status determination result is that the environment is abnormal, generate an internal item damage and environmental out-of-control hazard sign.

[0213] The internal item damage and environmental out-of-control hazard sign indicates a serious hazard when the items inside the transport container have undergone abnormal changes in condition and the internal temperature and humidity environment of the transport container exceeds the safe range.

[0214] When the internal item status is determined to be abnormal and the environmental status is determined to be abnormal, it indicates that the internal items of the transport container may be damaged, overturned, leaked or malfunction, and the temperature and humidity environment deviates from the safe storage requirements of the transport target. Therefore, a hazard sign for internal item damage and environmental out-of-control is generated.

[0215] S605: Generate a result level based on at least one of the following: regular bump condition indicator, enclosure damage hazard indicator, crushing deformation hazard indicator, internal movement hazard indicator, internal item damage and environmental out-of-control hazard indicator.

[0216] Based on the type and number of generated status identifiers, the corresponding result level is determined. When multiple status identifiers are generated simultaneously, the final result level is determined according to the priority order of severe risk level, high risk level, medium risk level, and low risk level, and the corresponding status identifier is retained as the alarm reason.

[0217] Generating internal movement hazard signs includes:

[0218] S700: When the internal item status determination result is that the solid target displacement is abnormal, and the environmental status determination result is that the environment is normal, a solid displacement hazard sign is generated.

[0219] Abnormal solid target displacement refers to a change in position, attitude, or trajectory of a solid target relative to its initial placement state that exceeds the permissible range. A solid displacement hazard sign is generated when an abnormal displacement of a solid target occurs inside a transport container, but the temperature and humidity environment inside the transport container has not exceeded the safe range.

[0220] When the internal item status determination result is that the solid target displacement is abnormal, and the environmental status determination result is that the environment is normal, it indicates that the solid target has undergone abnormal displacement, but has not yet caused the environmental status to be abnormal. Therefore, a solid displacement hazard label is generated.

[0221] S701: When the internal item status determination result is abnormal liquid flow and the environmental status determination result is normal, a hidden leakage hazard label is generated.

[0222] An abnormal flow of a liquid target refers to an abnormal change in the area of ​​the reflective region, continuous diffusion at the edges, an abnormal change in the flow direction, or a suspected leak. A hidden leak hazard sign is a hazard sign generated when abnormal flow or a suspected leak occurs in a liquid target inside a transfer container, but the temperature and humidity environment inside the transfer container has not exceeded the safe range.

[0223] When the internal item status assessment result indicates abnormal liquid flow and the environmental status assessment result indicates normal environment, it indicates that the liquid target has experienced abnormal flow or suspected leakage, but the abnormality has not yet caused the temperature and humidity environment to run out of control, thus generating a hidden leakage hazard sign.

[0224] S702: An internal movement hazard sign shall be generated when at least one of a solid displacement hazard sign and a hidden leakage hazard sign is present.

[0225] When at least one of a solid displacement hazard label and a hidden leakage hazard label is generated, it indicates that there is abnormal displacement, abnormal flow, or suspected leakage risk of the items inside the transfer container, and therefore an internal movement hazard label is generated.

[0226] The generated result levels include:

[0227] S800: When only a regular bump condition indicator is generated, the safety level assessment result is determined to be a low-risk level.

[0228] Low risk level refers to the risk level corresponding to when a transport container is subjected to a short-term impact or road bumps, but no abnormal deformation of the container, abnormal changes in the state of the internal items, or abnormal environment is detected.

[0229] When only a routine bump status indicator is generated, it indicates that the impact currently experienced by the transport container is a routine disturbance during transportation, and there is no accompanying damage to the container, abnormality of internal items, or abnormal temperature and humidity environment. Therefore, the safety level assessment result is determined to be a low-risk level.

[0230] S801: When an extrusion deformation hazard sign is generated, the safety level assessment result is determined to be a medium risk level.

[0231] The medium risk level refers to the risk level when no obvious transient impact was detected in the transport container, but the surface shape information of the container indicates that there is abnormal deformation of the container.

[0232] When a hazard sign for crushing deformation is generated, it indicates that the transport container may be subjected to continuous crushing, stacking, prying, or slow external force. Although no abnormalities have been detected in the internal items or the temperature and humidity environment, the structural condition of the container has already changed abnormally. Therefore, the safety level assessment result is determined to be medium risk.

[0233] S802: When a container damage hazard sign or an internal movement hazard sign is generated, the safety level assessment result is determined to be a high-risk level.

[0234] High-risk level refers to the risk level corresponding to when the transport container is subjected to abnormal impact and accompanied by damage to the container, or when the items inside the transport container have been abnormally displaced, flowed abnormally, or are suspected of leaking.

[0235] When a container damage hazard sign is generated, it indicates that the transport container has been subjected to an abnormal impact and that the outer surface of the container has undergone abnormal deformation. When an internal movement hazard sign is generated, it indicates that the items inside the transport container have experienced abnormal displacement, abnormal flow, or suspected leakage. These conditions may affect the safety of the transported item; therefore, the safety level assessment result is determined to be high-risk.

[0236] S803: When a hazard sign for internal item damage and environmental out-of-control occurs, the safety level assessment result is determined to be a severe risk level.

[0237] The severe risk level refers to the risk level corresponding to an abnormal change in the state of the items inside the transport container, and the temperature and humidity environment inside the transport container has exceeded the safe range.

[0238] When a hazard sign for internal item damage and environmental out-of-control is generated, it indicates that the items inside the transport container may be damaged, overturned, leaked, or become ineffective, and the temperature and humidity environment has deviated from the safe storage requirements of the transported object. There is a high risk that the transported object may be damaged, become ineffective, or cause a safety accident. Therefore, the safety level assessment result is determined to be a severe risk level.

[0239] S804: Generate corresponding result levels based on low risk level, medium risk level, high risk level, or severe risk level.

[0240] The corresponding risk level, categorized as low-risk, medium-risk, high-risk, or severe-risk, will be determined as the outcome level.

[0241] The alarm trigger information corresponding to the output intensity includes:

[0242] S900: When the safety level assessment result is low risk level, a status recording signal is generated, and the corresponding real-time inertial information, real-time temperature and humidity information and continuous video stream data are cached and recorded.

[0243] Status recording signals are signals used to record the current transportation status of transfer containers without triggering high-level alarm actions.

[0244] When the safety level assessment result is low risk, it indicates that the impact currently experienced by the transfer box is a normal transportation disturbance, and no abnormal deformation of the box body, abnormal changes in the state of the internal items, or environmental abnormalities are detected. Therefore, a status recording signal is generated, and real-time inertial information, real-time temperature and humidity information, and continuous video stream data for the corresponding time period are cached and recorded based on the status recording signal.

[0245] In this embodiment, when the safety decision tree does not generate any of the following: regular bump status indicator, enclosure damage hazard indicator, crush deformation hazard indicator, internal movement hazard indicator, and internal item damage and environmental out-of-control hazard indicator, the system continues to collect real-time inertial information, real-time temperature and humidity information, and continuous video stream data, and caches and records the corresponding data.

[0246] S901: When the safety level assessment result is medium risk level, a level 1 alarm signal is generated, and an abnormal prompt message is output based on the level 1 alarm signal, which is also used as alarm trigger information.

[0247] A Level 1 alarm signal is an alarm signal used to indicate that the transport container is in a medium-risk abnormal state.

[0248] When the safety level assessment result is medium risk level, it indicates that the transfer container may have abnormalities caused by squeezing deformation, stacking, prying or slow external force. Therefore, a level one alarm signal is generated, and abnormal prompt information is output according to the level one alarm signal to prompt the monitoring terminal or transportation personnel to pay attention to the status of the transfer container.

[0249] S902: When the safety level assessment result is high risk level, a level 2 alarm signal is generated, and the location of the abnormal area of ​​the cabinet or the abnormal area of ​​the internal items is determined based on the continuous video stream data corresponding to the safety level assessment result. At the same time, the location of the abnormal area of ​​the cabinet or the abnormal area of ​​the internal items is output according to the level 2 alarm signal as alarm trigger information.

[0250] Level 2 alarm signals are alarm signals used to indicate that the transport container is in a high-risk abnormal state.

[0251] When the safety level assessment result is high-risk, it indicates that the transport container may have damage to the container body, abnormal displacement of internal items, abnormal flow, or suspected leakage risks, thus generating a secondary alarm signal. Based on continuous video stream data, the location of the abnormal area within the container or the abnormal area within the internal items is determined, and this location is output according to the secondary alarm signal, enabling the monitoring terminal to pinpoint the location of the anomaly. The abnormal area location within the container refers to the image area location where damage, dents, edge displacement, or abnormalities in the enclosed structure of the container body are determined based on video frames of the outer surface. The abnormal area location within the internal items refers to the image area location where abnormal displacement of solid targets, abnormal flow of liquid targets, or suspected leakage are determined based on video frames of the internal item area.

[0252] S903: When the safety level assessment result is a severe risk level, a level 3 alarm signal is generated, and an alarm is executed according to the level 3 alarm signal. At the same time, based on continuous video stream data, images of abnormal areas where internal items are damaged or liquids are leaked are obtained, and images of abnormal areas, real-time inertial information, and real-time temperature and humidity information are uploaded as alarm trigger information.

[0253] Level 3 alarm signals are alarm signals used to indicate that the transport container is in a serious risk or abnormal state and to trigger an alarm action.

[0254] When the safety level assessment result is a severe risk level, it indicates that the items inside the transport container may have been damaged, overturned, leaked, or malfunctioned, and the temperature and humidity environment has deviated from the safe storage requirements of the transport target. Therefore, a level three alarm signal is generated. Local alarms and remote alarms are executed based on the level three alarm signal. At the same time, images of abnormal areas with corresponding internal item damage or liquid leakage are obtained based on continuous video stream data, and images of abnormal areas, real-time inertial information, and real-time temperature and humidity information are uploaded as alarm trigger information.

[0255] In this embodiment, when the transport target is a tuberculosis sample and the safety level assessment result is a severe risk level, in addition to triggering an alarm, biosafety event record information is generated, and abnormal area images, real-time inertial information, and real-time temperature and humidity information are uploaded for subsequent sample tracing, biosafety event handling, and emergency management.

[0256] Based on the same inventive concept, embodiments of the present invention provide a tuberculosis sample transport box monitoring and early warning system based on multi-source sensing, comprising:

[0257] The acquisition module is used to acquire real-time inertial information, real-time temperature and humidity information, and continuous video stream data;

[0258] The memory stores a program for implementing a control method for real-time monitoring and evaluation of the safety of a transport container based on multi-source sensing, as described in any one of the above descriptions.

[0259] The processor loads and executes programs stored in memory.

[0260] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0261] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A monitoring and early warning method for tuberculosis sample transport boxes based on multi-source sensing, characterized in that, include: Obtain the transportation target; Collect real-time inertial information, real-time temperature and humidity information, and continuous video stream data; Determine the container profile parameters based on the transportation objectives; The transient acceleration parameters are determined based on real-time inertial information and preset box inertial correction coefficients; By combining continuous video stream data and box contour parameters, information on the surface shape of the box and the status changes of the items inside are obtained; A safety decision tree is established based on the transportation objectives. Transient acceleration parameters, container surface shape information, internal item status change information, and real-time temperature and humidity information are input into the safety decision tree to generate a safety level assessment result. Output the corresponding alarm trigger information based on the security level assessment results.

2. The method for monitoring and early warning of tuberculosis sample transport boxes based on multi-source sensing according to claim 1, characterized in that, Obtaining information on the surface shape of the container and changes in the state of the items inside includes: Continuous video stream data includes video frames of the outer surface and video frames of the internal object area; Perform brightness normalization, filtering and noise reduction, and inter-frame stabilization on the external surface video frames to obtain preprocessed video frames; Based on the preset imaging correction parameters, distortion correction is performed on the preprocessed video frames to obtain the corrected video frames; Edge detection processing is performed on the corrected video frames to obtain the outer contour edge information of the box. The outer contour edge information of the box is matched and compared with the preset box contour parameters to determine the box edge offset, the box surface recessed area and the change of the closed structure. Generate the surface shape information of the box based on the offset of the box edge, the recessed area of ​​the box surface, and the change of the closed structure; Based on video frames of the internal object area and transient acceleration parameters, internal information is generated and used as information on changes in the state of the internal objects.

3. The method for monitoring and early warning of tuberculosis sample transport boxes based on multi-source sensing according to claim 2, characterized in that, The generated internal information includes: Transport targets include solid targets and / or liquid targets; The impact direction, impact intensity, and impact duration of the transfer box are determined based on transient acceleration parameters. When the transport target includes a solid target, target region identification is performed on the video frames of the internal item area to determine the solid detection area; Optical flow estimation is performed on continuous solid detection regions to determine the pixel motion direction and pixel displacement of solid targets between adjacent video frames; The pixel motion direction and pixel displacement are matched and analyzed with the impact direction, impact intensity and impact duration to generate solid target displacement information; When the transport target includes a liquid target, determine the abnormal liquid flow information based on the liquid target; Internal information is generated based on at least one of the abnormal liquid flow information and solid target displacement information.

4. The method for monitoring and early warning of tuberculosis sample transport boxes based on multi-source sensing according to claim 3, characterized in that, Determine abnormal liquid flow information based on the liquid target, including: Determine the visible area of ​​the liquid based on the liquid target; Based on the liquid visible area information, target area identification is performed on video frames of the internal object area to determine the liquid detection area; Extract the changes in the area of ​​the liquid reflective region, the liquid edge diffusion velocity, and the changes in the liquid flow direction from the liquid detection area; By matching and analyzing the changes in the area of ​​the liquid reflective region, the liquid edge diffusion velocity, and the changes in the liquid flow direction with the impact intensity, impact direction, and impact duration, abnormal liquid flow information is generated.

5. A monitoring and early warning method for tuberculosis sample transport boxes based on multi-source sensing according to claim 3, characterized in that, The generated security level assessment results include: Based on the transportation objectives, the impact threshold, container deformation threshold, internal state change threshold, and temperature and humidity safety range are determined. The transient acceleration parameters are compared with the impact threshold to generate an impact state determination result. The surface shape information of the enclosure is compared with the deformation threshold of the enclosure to generate the deformation state determination result of the enclosure; The internal item status change information is compared with the internal status change threshold to generate an internal item status determination result. The real-time temperature and humidity information is compared with the safe temperature and humidity range to generate an environmental condition judgment result. The impact condition determination results, enclosure deformation condition determination results, internal item condition determination results, and environmental condition determination results are input into the safety decision tree to generate a result level as the safety level assessment result.

6. A monitoring and early warning method for tuberculosis sample transport boxes based on multi-source sensing according to claim 5, characterized in that, The generated result levels include: When the impact condition determination result is an impact anomaly and the box deformation condition determination result is a box deformation normal, a regular bump condition indicator is generated. When the impact condition determination result is impact abnormal and the box deformation condition determination result is box deformation abnormal, a box damage hazard sign is generated. When the impact condition determination result is normal and the box deformation condition determination result is abnormal, a crush deformation hazard sign is generated. When the internal item status is determined to be abnormal and the environmental status is determined to be normal, an internal movement hazard sign is generated. When the internal item status is determined to be abnormal and the environmental status is determined to be abnormal, a hazard sign for internal item damage and environmental out-of-control is generated. The result level is generated based on at least one of the following: conventional bump condition warning, container damage hazard warning, crushing deformation hazard warning, internal movement hazard warning, internal item damage and environmental out-of-control hazard warning.

7. A monitoring and early warning method for tuberculosis sample transport boxes based on multi-source sensing according to claim 6, characterized in that, Generating internal movement hazard signs includes: When the internal item status assessment result is that the solid target displacement is abnormal, and the environmental status assessment result is that the environment is normal, a solid displacement hazard sign is generated. When the internal item status assessment result indicates abnormal liquid flow and the environmental status assessment result indicates normal environment, a hidden leak hazard sign is generated. An internal movement hazard sign is generated when at least one of a solid displacement hazard sign and a hidden leakage hazard sign is present.

8. A monitoring and early warning method for tuberculosis sample transport boxes based on multi-source sensing according to claim 6, characterized in that, The generated result levels include: When only a regular bump condition indicator is generated, the safety level assessment result is determined to be low risk. When an extrusion deformation hazard sign is generated, the safety level assessment result is determined to be of medium risk level; When a hazard sign for damage to the enclosure or an hazard sign for internal movement is generated, the safety level assessment result is determined to be a high-risk level. When a hazard sign indicating damage to internal items and loss of environmental control is generated, the safety assessment result is determined to be a severe risk level. The corresponding result level is generated based on the low risk level, medium risk level, high risk level, or severe risk level.

9. A monitoring and early warning method for tuberculosis sample transport boxes based on multi-source sensing according to claim 8, characterized in that, The alarm trigger information corresponding to the output intensity includes: When the safety level assessment result is low risk, a status recording signal is generated, and the corresponding real-time inertial information, real-time temperature and humidity information, and continuous video stream data are cached and recorded. When the safety level assessment result is medium risk level, a level 1 alarm signal is generated, and an abnormal prompt message is output based on the level 1 alarm signal, which is also used as the alarm trigger information. When the safety level assessment result is high risk level, a level 2 alarm signal is generated, and the location of the abnormal area of ​​the cabinet or the abnormal area of ​​the internal items is determined based on continuous video stream data, and the location of the abnormal area of ​​the cabinet or the abnormal area of ​​the internal items is output according to the level 2 alarm signal as alarm trigger information. When the safety level assessment result is a severe risk level, a level 3 alarm signal is generated and an alarm is executed based on the level 3 alarm signal. At the same time, images of abnormal areas with corresponding internal item damage or liquid leakage are obtained based on continuous video stream data, and abnormal area images, real-time inertial information, and real-time temperature and humidity information are uploaded as alarm trigger information.

10. A monitoring and early warning system for tuberculosis sample transport boxes based on multi-source sensing, characterized in that, include: The acquisition module is used to acquire real-time inertial information, real-time temperature and humidity information, and continuous video stream data; The memory stores a program for implementing a control method for real-time monitoring and evaluation of the safety of a transport container based on multi-source sensing, as described in any one of claims 1 to 9. The processor loads and executes programs stored in memory.

Images