Industrial safety control method and device, computer device and storage medium

Abstract

This invention discloses an industrial safety control method and related equipment. The method includes: acquiring depth data of a monitoring area; performing interference identification processing on the depth data to determine valid and invalid depth data, and determining the observation spatial range of the pixels corresponding to the invalid depth data; based on the intersection relationship between the observation spatial range and a preset protection zone, determining unknown occupied areas within the preset protection zone where the presence of targets cannot be ruled out; determining the occupancy status of the preset protection zone based on the valid depth data and the unknown occupied areas; and outputting a corresponding industrial safety control signal based on the occupancy status. By determining the unknown occupied areas where the presence of targets cannot be ruled out through the intersection relationship between the observation spatial range of the pixels corresponding to the invalid depth data and the preset protection zone, and by including the unknown occupied areas in the occupancy judgment of the protection zone based on the presence of targets, the reliability of the industrial safety control signal output can be effectively improved.

Description

Technical Field

[0001] This invention relates to the field of industrial safety, and more particularly to an industrial safety control method, apparatus, computer equipment, and storage medium. Background Technology

[0002] With the development of industrial automation technology, industrial safety protection is gradually evolving from fixed fences and contact detection to non-contact spatial monitoring. In scenarios such as robot workstations, automated production lines, and logistics conveying equipment, depth-sensing-based industrial safety monitoring equipment can acquire depth data of the monitored area, determine the spatial location of personnel or objects, and judge whether personnel or objects have entered a pre-set protected area. It can then output safety control signals such as stopping operation and limiting speed to the controlled equipment, reducing the risk of collisions, crushing accidents, and other incidents when personnel approach dangerous equipment.

[0003] In real-world industrial environments, strong reflections from metal surfaces, low-reflectivity materials, welding arc light, dust and fumes, and optical interference between multiple depth sensors can all affect the reliability of depth data, leading to measurement deviations, numerical anomalies, or missing depth values ​​in local pixels. Existing depth data processing methods typically improve measurement results through filtering, outlier removal, or data repair, and use the processed depth data to determine the presence of targets within the protected area. For industrial safety, target detection results directly affect whether hazardous equipment can continue to operate. Therefore, abnormal depth data not only affects the accuracy of target location measurement but may also hinder the timely triggering of protective measures when personnel enter hazardous areas.

[0004] However, after abnormal depth data is removed, the corresponding locations may not participate in target detection due to a lack of effective measurement results; the repaired data may also deviate from the actual scene. In this case, relying solely on the processed depth data for target detection can easily lead to the misjudgment that no target is detected as the absence of a target within the protected area, resulting in the missed detection of real targets within the affected area. This allows controlled equipment to continue dangerous activities even after personnel have entered the protected area, thereby affecting the reliability of industrial safety protection.

[0005] Therefore, how to improve the reliability of industrial safety control signal output when depth data is affected by industrial environmental interference is a technical problem that needs to be solved. Summary of the Invention

[0006] Therefore, it is necessary to provide an industrial safety control method, device, computer equipment, and storage medium to address the above-mentioned technical problems, which has the advantage of improving the reliability of industrial safety control signal output when deep data is interfered with by the industrial environment.

[0007] An industrial safety control method, the method comprising: Acquire depth data of the monitoring area, which includes a preset protected area; The depth data is subjected to interference identification processing to determine the valid depth data and invalid depth data in the depth data, and the observation space range of the pixels corresponding to the invalid depth data is determined. Based on the intersection of the observation space range and the preset protection zone, unknown occupied areas within the preset protection zone that cannot be ruled out as having targets are determined; Based on the effective depth data and the unknown occupied area, the occupancy status of the preset protected area is determined, wherein the unknown occupied area is judged to be occupied based on the presence of a target; Based on the occupancy status, the corresponding industrial safety control signal is output.

[0008] Optionally, the step of performing interference identification processing on the depth data to determine the valid and invalid depth data includes: Obtain optical measurement information corresponding to the depth data, and determine interference features based on the optical measurement information and the depth data. The interference features include at least two of the following: echo intensity features, ambient light features, saturation features, cross-frequency consistency features, and cross-frame consistency features. Based on the interference characteristics, the interference type and degree of interference for the corresponding pixel are determined; Based on the type and degree of interference, the validity of the depth data is determined to identify the valid depth data and the invalid depth data.

[0009] Optionally, determining the observation spatial range of the pixels corresponding to the invalid depth data includes: Based on the spatial distribution of pixels corresponding to the invalid depth data, the abnormal pixel region is determined; According to a preset expansion rule, the abnormal pixel region is expanded to obtain the target pixel region; The spatial range corresponding to the target pixel region is determined based on the observation direction corresponding to the target pixel region; The portion of the spatial range that falls within a preset distance range is defined as the observation spatial range.

[0010] Optionally, determining the unknown occupied area within the preset protection zone where the presence of a target cannot be ruled out, based on the intersection relationship between the observation spatial range and the preset protection zone, includes: Determine the area where the observation space range intersects with the preset protection zone; Based on the spatial range of the intersecting region and the preset minimum target size, determine whether the intersecting region can accommodate the preset minimum target; Intersecting regions that cannot accommodate the preset minimum target are excluded from the unknown occupied regions; The unknown occupied area is determined based on the intersecting area that can accommodate the preset minimum target.

[0011] Optionally, determining the occupancy status of the preset protected area based on the effective depth data and the unknown occupied area includes: Based on the result of the interference identification process, the target depth data that has been interfered with is determined from the effective depth data, and the distance error range corresponding to the target depth data is determined. Based on the distance error range, the target location range corresponding to the target depth data is expanded; Determine whether the expanded target location range intersects with the preset protected area to obtain the intersection determination result; Based on the intersection judgment result, the effective depth data other than the target depth data, and the unknown occupied area, the occupancy status of the preset protected area is determined.

[0012] Optionally, the step of outputting a corresponding industrial safety control signal based on the occupancy status includes: Based on the effective depth data and the regional distribution of the invalid depth data in multiple consecutive acquisition cycles, regional risk information is determined. The regional risk information includes the effective coverage rate of the preset key area within the preset protection zone, the spatial range of the invalid connected area, and the duration of interference. Based on the type of interference, the degree of interference, and the regional risk information, the industrial safety control level is determined, and the corresponding acquisition parameters and protection parameters are determined based on the industrial safety control level. The acquisition parameters include modulation frequency and exposure parameters, and the protection parameters include effective monitoring range, adjustment range of the preset protection zone, and allowable operating speed of the equipment. When the occupancy status indicates that the preset protected area is occupied, or when the area risk information meets the preset stop conditions, a safety stop signal is output; When the occupancy status indicates that the preset protected area is vacant and the area risk information does not meet the preset stop condition, monitoring is performed according to the acquisition parameters and the protection parameters, and the corresponding industrial safety control signal is output according to the re-determined occupancy status.

[0013] Optionally, the method further includes: When the risk information in the area meets the entry conditions corresponding to a higher industrial security control level, the current industrial security control level will be switched to the higher industrial security control level. When the risk information in the area meets the exit conditions corresponding to the current industrial safety control level, continuous verification is performed based on the re-collected depth data. The risk level required by the exit conditions is lower than the risk level when entering the current industrial safety control level. If, within a consecutive preset number of verification cycles, the effective coverage, cross-frequency consistency, and occupancy status of the key area all meet the preset recovery conditions, and the self-test results are normal and the occupancy status of the preset protected area determined by the two processing channels is consistent, the current industrial safety control level is reduced or the operation permit is restored. During the continuous verification, it is prohibited to use historical depth data prior to entering the current industrial safety control level to determine that the preset protected area is in an idle state; When the duration of the interference exceeds the preset time limit or the number of recovery failures reaches the preset number, the latching fault state is entered. After the preset manual reset conditions are met, the fault latch is released and the continuous verification is re-executed, and the safety stop signal is maintained until the preset recovery conditions are met.

[0014] An industrial safety control device, the device comprising: The acquisition module is used to acquire depth data of the monitoring area, which includes a preset protected area; The identification module is used to perform interference identification processing on the depth data, determine the valid depth data and invalid depth data in the depth data, and determine the observation space range of the pixels corresponding to the invalid depth data. The first determining module is used to determine, based on the intersection relationship between the observation spatial range and the preset protection zone, the unknown occupied area within the preset protection zone where the presence of a target cannot be ruled out; The second determining module is used to determine the occupancy status of the preset protected area based on the effective depth data and the unknown occupancy area, wherein the unknown occupancy area is determined to be occupied based on the presence of a target. The output module is used to output the corresponding industrial safety control signal based on the occupancy status.

[0015] A computer device includes a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, wherein the processor implements the aforementioned industrial safety control method when executing the computer-readable instructions.

[0016] A readable storage medium having computer-readable instructions stored thereon, which, when executed by a processor, implement the aforementioned industrial safety control method.

[0017] The aforementioned industrial safety control method, apparatus, computer equipment, and storage medium acquire depth data of a monitoring area, which includes a preset protection zone; perform interference identification processing on the depth data to determine valid and invalid depth data, and determine the observation spatial range of the pixels corresponding to the invalid depth data; based on the intersection relationship between the observation spatial range and the preset protection zone, determine unknown occupied areas within the preset protection zone where the presence of a target cannot be ruled out; based on the valid depth data and the unknown occupied areas, determine the occupancy status of the preset protection zone, wherein the unknown occupied areas are judged as having a target; based on the occupancy status, output a corresponding industrial safety control signal. By performing interference identification on the depth data, distinguishing between valid and invalid depth data, and determining unknown occupied areas where the presence of a target cannot be ruled out based on the intersection relationship between the observation spatial range of the pixels corresponding to the invalid depth data and the preset protection zone, the unknown occupied areas are judged as having a target and are included in the occupancy judgment of the protection zone. This avoids misjudging corresponding areas as empty areas due to invalid depth data, reduces the risk of missed target detection under interference conditions, and improves the reliability of industrial safety control signal output. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating an industrial safety control method according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating another industrial safety control method according to one embodiment of the present invention; Figure 3 This is a schematic diagram of a process for determining the occupancy status of a preset protected area in one embodiment of the present invention; Figure 4 This is a schematic diagram of industrial safety control level switching in one embodiment of the present invention; Figure 5 This is a schematic diagram of an industrial safety control device according to one embodiment of the present invention; Figure 6 This is a schematic diagram of a computer device according to an embodiment of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] In one embodiment, such as Figure 1 As shown, an industrial safety control method is provided, comprising the following steps: 101. Obtain depth data of the monitoring area.

[0022] In this embodiment of the invention, the industrial safety control method can be applied to electronic devices with depth data processing and safety signal output functions, such as three-dimensional safety sensors, industrial safety controllers or industrial computers that are communicatively connected to the depth sensors. The electronic devices can process the collected depth data and output industrial safety control signals to the controlled devices, which may include industrial robots, rolling mills, die-cutting equipment or logistics conveying equipment.

[0023] The monitoring area can be the spatial region covered by the field of view and monitoring range of the depth sensor, such as the space around a robot workstation, the space on the feed side of a roller press, or the personnel passage space near a logistics conveyor line. The monitoring area includes a preset protection zone and may also include a warning zone located outside the preset protection zone to meet the needs of hazard proximity judgment and early warning, respectively.

[0024] A preset protection zone is a three-dimensional spatial area pre-defined based on the hazardous location, range of motion, and distance required to stop the hazardous movement of the controlled equipment. It is used to determine whether personnel, body parts, or other targets needing detection have entered the area requiring safety protection measures. The preset protection zone can be represented by spatial boundary coordinates, and its shape can be a cuboid, polyhedron, or an irregular spatial shape adapted to the hazardous area of ​​the equipment. For example, a preset protection zone can be set on the feed side of a roller press so that personnel can enter the preset protection zone when their hands approach the hazardous roller. The location and boundaries of the preset protection zone can be pre-stored and use a unified coordinate system with the spatial position obtained by converting depth data.

[0025] Depth data can represent the distance between the surface of an object within the monitoring area and the depth sensor. Specifically, it can be represented as a depth map, which can be a single-frame depth map or a sequence of multiple depth maps. Pixels in the depth map correspond to the observation direction of the depth sensor, and the depth value of a pixel represents the distance measurement result obtained along the corresponding observation direction. For pixels for which no usable distance measurement result was obtained, the corresponding pixel location and invalid value marker can be retained for subsequent interference identification and observation space range determination, avoiding the direct loss of information from the corresponding pixels during data acquisition.

[0026] Specifically, depth data can be acquired using a time-of-flight depth sensor. The time-of-flight depth sensor emits modulated light into the monitoring area and receives the echoes reflected from the object's surface. Based on the phase difference between the echo and a reference signal, the corresponding distance is calculated, forming a depth map. Electronic devices can read the depth map output by the built-in depth sensor or receive depth maps sent by an external depth sensor via a communication interface, retaining the correspondence between each pixel and its depth value. Depth data can be continuously acquired according to a preset acquisition cycle to support ongoing safety monitoring of the monitored area.

[0027] 102. Perform interference identification processing on the depth data to determine the valid and invalid depth data, and determine the observation space range of the pixels corresponding to the invalid depth data.

[0028] In this embodiment of the invention, interference identification processing is used to determine whether depth data is affected by external lighting, object reflection characteristics, or other ranging devices, and whether the affected depth data can still be used for target location determination. Interference can include strong ambient light interference, saturation interference, low echo interference, multipath interference, multi-machine crosstalk, and scattering interference caused by dust and smoke. For example, strong reflection from a metal roller may cause local pixel saturation, black materials may cause insufficient echo intensity, and light emitted by other depth sensors may cause abnormal fluctuations in the measured distance.

[0029] Specifically, depth data of corresponding pixels at different modulation frequencies and depth changes of corresponding pixels over multiple consecutive acquisition cycles can be compared to determine if there are any anomalies inconsistent with normal measurements or target motion. Interference identification can also be processed by combining echo intensity, ambient light intensity, and saturation indicators corresponding to the depth data. For example, if the depth value of the corresponding pixel shows significant differences at different modulation frequencies when the target position remains stable, accompanied by abnormal echoes, it can be determined that the corresponding depth data is interfered with. For moving targets, the spatial distribution of adjacent pixels and the continuity of target motion can be considered to avoid identifying normal motion as interference simply because the depth value changes over time.

[0030] Effective depth data refers to depth data that meets preset validity conditions and can be used for target location determination. Preset validity conditions can be set based on the measurement performance of the depth sensor and safety monitoring requirements. For example, the echo intensity must meet distance calculation requirements, pixels must not be saturated, and measurement results from different modulation frequencies or continuous acquisition periods must meet corresponding consistency requirements. Effective depth data can include both stable and reliable measurement data and data that, despite some interference, still provides usable distance information; the latter can be processed in subsequent occupancy determination based on the distance error range.

[0031] Invalid depth data refers to depth data that does not meet preset validity conditions and cannot provide reliable target distance. It can include data generated due to saturation, weak echo, distance calculation failure, or severe measurement anomalies. Invalid depth data may manifest as missing values, invalid value markers, or even as measured values ​​with normal numerical form but inconsistent with the actual distance. Therefore, validity cannot be determined solely based on the presence of depth values. Electronic devices can associate validity determination results with corresponding depth values ​​for storage, retaining the pixel location and invalidity determination result for invalid depth data, and prohibiting the direct use of invalid depth values ​​to prove that the corresponding space is free.

[0032] The observation space range of a pixel refers to the three-dimensional spatial range determined by the pixel's imaging position, observation direction, and preset distance range. Although invalid depth data cannot provide reliable distance, the position of the corresponding pixel in the photosensitive array is still determined. Therefore, the corresponding observation direction and angular coverage range can be determined based on the calibration parameters of the depth sensor. Starting from the optical center of the depth sensor and extending outward along the pixel's angular coverage range, a corresponding frustum can be formed. Limiting the frustum between a preset minimum reliable distance and a preset maximum safe range yields the pixel's observation space range. The determination of the observation space range does not depend on the distance indicated by the invalid depth value.

[0033] When multiple invalid pixels are distributed adjacently in the depth map, the region formed by adjacent invalid pixels can be mapped to three-dimensional space to determine a common observation space range. For areas that may be affected by optical scattering, neighbor crosstalk, or calibration errors, the pixel region can be appropriately expanded before determining the corresponding observation space range. For example, if some pixels near black material cannot be reliably measured due to weak echoes, the corresponding observation direction can still be determined based on the pixel position, and the spatial range extending along the observation direction can be determined for subsequent judgment of the intersection relationship with the preset protection zone.

[0034] 103. Based on the intersection of the observation space range and the preset protection zone, determine the unknown occupied areas within the preset protection zone where the presence of targets cannot be ruled out.

[0035] In this embodiment of the invention, the observation space range represents the spatial range where the distance measurement of invalid pixels cannot be reliably confirmed, and the preset protection zone represents the three-dimensional spatial region where target occupancy judgment needs to be performed. By placing the observation space range and the preset protection zone in the same coordinate system, it is possible to determine whether the two intersect based on their respective spatial boundaries, and if they intersect, determine the position and range of the intersecting part. If the two do not intersect, the corresponding invalid pixels are not used to form the unknown occupancy area within the preset protection zone; if the two intersect, it is necessary to further determine whether the intersecting part can exclude the existence of a target.

[0036] An unknown occupied area refers to a region within a pre-defined protected area where the presence of a target cannot be confirmed due to invalid depth data. Determining an unknown occupied area does not require the detection of a target, nor does it indicate that the target's precise location has been obtained. Rather, it indicates that existing measurement results are insufficient to support the conclusion that the corresponding area is free. Therefore, the absence of a target in an intersecting region cannot be automatically assumed simply because there is no corresponding valid point cloud data or invalid depth data has been removed.

[0037] Specifically, the intersecting area can be calculated based on the observation space range and the boundary of the preset protected area. Intersecting areas where the presence of a target cannot be ruled out through effective measurement are marked as unknown occupied areas. In one implementation, the intersecting area can be directly identified as an unknown occupied area. In another implementation, a preset minimum target size can be used for screening; when the intersecting area can accommodate the minimum target to be detected, the target may be completely within the range that cannot be reliably measured, so the corresponding area is identified as an unknown occupied area. For intersecting areas that cannot accommodate the preset minimum target, if effective observations in the surrounding area are sufficient to detect the rest of the target, they can be excluded from the unknown occupied area. The preset minimum target size is determined according to the personnel, body parts, or other targets to be detected, and the screening results do not affect the identification of the real target based on effective depth data.

[0038] For example, on the feed side of the roller press, some pixels are deemed invalid due to weak echoes. If the corresponding observation space extends beyond a preset protection zone, the intersecting portion within the preset protection zone is considered the area to be judged. When the intersecting portion can accommodate the hand target to be detected, and reliable distance information is lacking, even if no valid measurement point corresponding to the hand is formed in the depth map, the existence of the hand cannot be ruled out, and the intersecting portion should be determined as an unknown occupied area. The portion of the observation space outside the preset protection zone is not included in the unknown occupied area of ​​this protection zone.

[0039] For observation spatial ranges formed by multiple invalid pixels, the intersections with preset protected areas can be calculated separately, and the overlapping parts can be merged to form unknown occupied areas for subsequent occupancy determination. This preserves the spatial impact corresponding to the failure of depth data, allowing protected areas that cannot be reliably observed to continue participating in safety assessments.

[0040] 104. Based on the effective depth data and the unknown occupied area, determine the occupancy status of the preset protected area.

[0041] In this embodiment of the invention, the occupancy status is used to indicate whether there is a target in the preset protected area that requires safety protection measures. The occupancy status can include an occupied status and an idle status. When determining the occupancy status, the target location reflected by the effective depth data and the location and range of the unknown occupied area are considered simultaneously, so that both the actually detected targets and the areas where the presence of targets cannot be ruled out are involved in the judgment.

[0042] For valid depth data, the spatial location of the target can be determined based on the observation direction and measurement distance corresponding to each pixel, and then transformed into the coordinate system corresponding to the preset protection zone. By comparing the spatial location of the target with the boundary of the preset protection zone, it is determined whether the target has entered the preset protection zone. When the target location corresponding to the valid depth data falls within the preset protection zone, the preset protection zone is considered occupied. For depth data that is disturbed but still valid, the possible location range of the target can be determined by combining the distance error range; when the possible location range of the target intersects with the preset protection zone, it is also considered occupied as if a target is present.

[0043] For unknown occupied areas, since reliable distance measurements are lacking at the corresponding locations, the existence of a target cannot be confirmed. Therefore, the same processing result as for existing targets is applied in the occupancy determination. As long as a confirmed unknown occupied area exists within the preset protected area, the preset protected area will be determined as occupied, even if the effective depth data does not reflect that the target has entered the preset protected area. Occupancy determination based on existing targets does not require generating fictitious target distances or target outlines for unknown occupied areas; this can be achieved by retaining area markers and having the area markers participate in the occupancy calculation.

[0044] Specifically, when the effective depth data indicates the presence of a target within the pre-defined protected area, or when there is an unknown occupied area within the pre-defined protected area, the pre-defined protected area is determined to be occupied. When the effective depth data does not indicate the presence of a target, there is no unknown occupied area, and the pre-defined monitoring coverage requirements are met, the pre-defined protected area is determined to be vacant. A vacant status cannot be determined solely because no measurement point falls within the pre-defined protected area in the effective depth data.

[0045] For example, in a robotic workstation, the target locations corresponding to the effective depth data are all outside the preset protection zone. However, welding arc light causes some pixels to saturate, and the corresponding observation space range forms an unknown occupied area within the preset protection zone. In this case, the preset protection zone is still classified as occupied to avoid missing personnel who may enter the protection zone due to local measurement failure. In cases where both actual target occupancy and unknown occupancy exist simultaneously, either occupancy criterion is sufficient to maintain the preset protection zone's occupied state, and it is not negated by effective idle measurement results in another area.

[0046] 105. Based on the occupancy status, output the corresponding industrial safety control signal.

[0047] In this embodiment of the invention, industrial safety control signals are used to transmit control commands to the controlled equipment corresponding to the occupancy status of a preset protected area, in order to control the dangerous movements of the controlled equipment. Industrial safety control signals may include safety stop signals and operating permission signals, and can be sent to the controller of the controlled equipment through a safety output interface or a safety communication interface.

[0048] When the preset protection zone is occupied, the electronic equipment outputs a safety stop signal, revoking the operating permission of the controlled equipment and causing it to stop or prevent the initiation of any dangerous movements. Whether the occupancy is triggered by a target detected by effective depth data or by an unknown occupancy area, the corresponding safety stop control is executed. For example, if a hand is detected entering the preset protection zone on the feed side of the roller press, or if the hand cannot be ruled out as being within the preset protection zone due to local depth data failure, a safety stop signal can be output to control the dangerous roller to stop rotating.

[0049] When the preset protected area is idle and there are no faults or other safety restrictions preventing the output of the operation permit, the electronic equipment can output an operation permit signal. The operation permit signal indicates that the current monitoring results allow the controlled equipment to operate. Whether the controlled equipment starts can also be determined by the equipment controller in combination with the start command and other interlocking conditions, avoiding equating the restoration of the protected area to idleness directly with the automatic start of the equipment.

[0050] Industrial safety control signals can be output via a dual-channel approach, transmitting the corresponding safety status through two output channels. In the implementation using dual processing channels, the occupancy status of preset protected areas can be calculated separately, and the calculation results can be compared. When the results from the two processing channels are inconsistent, processing times out, or communication data verification anomalies occur, a safety stop signal is output. The generation of industrial safety control signals can be independent of the ordinary point cloud output used for display, avoiding the impact of ordinary point cloud publishing or display communication anomalies on the output of the safety stop signal.

[0051] During continuous monitoring, the electronic equipment updates the industrial safety control signals based on the occupancy status determined in each acquisition cycle. When the preset protection zone changes from idle to occupied, the operation permit is revoked in a timely manner; when the preset protection zone changes from occupied to idle, the operation permit signal can be re-output after continuous verification and corresponding recovery conditions are met, preventing the controlled equipment from repeatedly starting and stopping due to instantaneous measurement recovery.

[0052] In this embodiment of the invention, depth data of a monitoring area, including a preset protection zone, is acquired. Interference identification processing is performed on the depth data to determine valid and invalid depth data, and the observation spatial range of the pixels corresponding to the invalid depth data is determined. Based on the intersection of the observation spatial range and the preset protection zone, unknown occupied areas within the preset protection zone where the presence of a target cannot be ruled out are identified. Based on the valid depth data and the unknown occupied areas, the occupancy status of the preset protection zone is determined, wherein the unknown occupied areas are judged as having a target. Based on the occupancy status, a corresponding industrial safety control signal is output. By performing interference identification on the depth data, distinguishing between valid and invalid depth data, and determining unknown occupied areas where the presence of a target cannot be ruled out based on the intersection of the observation spatial range of the pixels corresponding to the invalid depth data and the preset protection zone, the unknown occupied areas are judged as having a target in the protection zone occupancy determination. This avoids misjudging corresponding areas as empty areas due to invalid depth data, reduces the risk of missed target detection under interference conditions, and improves the reliability of the industrial safety control signal output.

[0053] Optionally, in the step of performing interference identification processing on the depth data to determine the valid and invalid depth data, optical measurement information corresponding to the depth data can also be obtained, and interference features can be determined based on the optical measurement information and the depth data. The interference features include at least two of the following: echo intensity features, ambient light features, saturation features, cross-frequency consistency features, and cross-frame consistency features. Based on the interference features, the interference type and degree of the corresponding pixel are determined. Based on the interference type and degree of interference, the validity of the depth data is determined to identify the valid and invalid depth data.

[0054] In this embodiment of the invention, optical measurement information refers to data obtained by the depth sensor during acquisition and distance calculation that reflects the state of the received optical signal. This data may include modulation amplitude, ambient light estimate, phase sampling value, and saturation flag. The optical measurement information is correlated with depth data according to pixel position, acquisition time, and modulation frequency, and is used to analyze the causes and impacts of depth measurement anomalies. Interference characteristics are abnormal behaviors extracted from the optical measurement information and depth data. Interference type indicates the category of interference corresponding to the abnormal behavior, and interference degree indicates the magnitude of the interference's impact on the reliability of distance measurement.

[0055] Echo intensity characteristics can be determined based on modulation amplitude and noise level. Specifically, the modulation amplitude can be compared with a preset minimum amplitude, or the ratio of modulation amplitude to noise level can be calculated to determine whether there is sufficient effective echo for distance calculation. Ambient light characteristics can be determined by background data collected with active lighting off, or estimated based on common components in different phase sample values, and combined with continuous acquisition results to determine the amplitude of ambient light variation. Saturation characteristics can be determined based on whether the sampled values ​​reach the full scale of the analog-to-digital converter and whether clipping occurs in the phase sample values, reflecting whether the received signal exceeds the normal measurable range.

[0056] Cross-frequency consistency features reflect the degree of agreement between measurement results at different modulation frequencies. This can be determined by comparing the corresponding pixel depth values ​​after distance ambiguity resolution. For example, the difference between depth values ​​at different frequencies is calculated and compared with a preset allowable deviation. If the difference exceeds the allowable deviation, it indicates a cross-frequency measurement anomaly. Cross-frame consistency features reflect the stability of continuous acquisition results. This can be determined by statistically analyzing depth fluctuations, phase jump counts, or echo amplitude changes within a preset time window. For moving targets, motion estimation can be combined with comparisons of the deviation between the measurement results and the expected position to avoid directly classifying depth changes caused by normal motion as interference.

[0057] When determining interference types based on interference characteristics, feature combinations and judgment conditions corresponding to different interference types can be preset. For example, if the ambient light estimate is high or changes rapidly, accompanied by saturation characteristics, strong ambient light interference can be determined; if the echo intensity is lower than the preset requirement and the depth value is unstable, low echo interference can be determined; if cross-frequency measurement results are inconsistent, accompanied by neighborhood phase anomalies near reflective surfaces, multipath interference can be determined; if abnormal jumps in depth or phase occur across frames, accompanied by large-scale synchronization fringes, multi-machine crosstalk can be determined. When the same pixel meets the judgment conditions for multiple types of interference, multiple interference types can be retained.

[0058] The degree of interference can be determined based on the magnitude of the deviation of the interference features from the normal measurement range and the duration of the abnormality. Specifically, the amount of insufficient echo intensity, excessive ambient light, cross-frequency depth difference, and cross-frame fluctuation amplitude can be compared with pre-calibrated classification thresholds to obtain the interference level corresponding to each feature. Then, the overall interference level of the pixel is determined according to a preset combination rule. For example, the highest level among the various interference levels can be taken as the overall interference level; alternatively, each feature can be converted into an evaluation value and weighted, and the interference can be classified into mild, moderate, and severe based on the weighting result. The classification thresholds and combination rules are determined based on the calibration relationship between measurement error and interference features.

[0059] When determining the validity of data based on the type and severity of interference, first check whether the corresponding interference type triggers the conditions for direct invalidation. For example, if saturation interference causes phase sampling value clipping, or low echo interference causes the modulation amplitude to be lower than the distance calculation requirement, the corresponding depth data will be determined as invalid; if phase calculation fails, it will also be determined as invalid. For data that does not trigger the direct invalidation condition, then determine whether the measurement error is within the allowable range based on the severity of interference. Data with mild interference that meets normal measurement accuracy requirements is determined as valid depth data; data subject to some interference, but whose distance error can still be limited within a preset range, can also be determined as valid depth data, and the corresponding distance error range will be considered in subsequent occupancy determinations; data with severe interference that cannot reliably limit the distance error is determined as invalid depth data.

[0060] For example, in a scenario where highly reflective rollers coexist with black materials, if the pixel corresponding to the roller experiences sampling saturation, it should be deemed invalid even if it outputs a depth value in a normal numerical form. If the pixel corresponding to the black material has weak echoes but still meets the distance calculation requirements and the cross-frequency and cross-frame measurement deviations are within the allowable range, it can be retained as valid depth data; if the echoes are consistently insufficient and accompanied by obvious measurement anomalies, it should be deemed invalid. The determination results are saved in association with the pixel positions for subsequent target location determination and observation space range determination.

[0061] Optionally, in the step of determining the observation spatial range of the pixels corresponding to invalid depth data, the abnormal pixel region can be determined according to the spatial distribution of the pixels corresponding to invalid depth data; the abnormal pixel region can be expanded according to a preset expansion rule to obtain the target pixel region; the spatial range corresponding to the target pixel region can be determined according to the observation direction corresponding to the target pixel region; and the part of the spatial range located within a preset distance range can be determined as the observation spatial range.

[0062] In this embodiment of the invention, abnormal pixel regions can be obtained by performing connected component analysis on the pixels corresponding to invalid depth data. Specifically, based on the row and column positions of the pixels in the depth map, invalid pixels that are adjacent at edges or corners are grouped into the same connected region, and the pixel set and boundary of each connected region are recorded as abnormal pixel regions. For mutually separated invalid pixels, their corresponding regions can be retained separately, avoiding the direct merging of large areas with reliable measurement results into abnormal regions.

[0063] The preset expansion rule is used to determine the extent to which abnormal pixel regions expand outwards, including pixels potentially affected by neighboring crosstalk, trailing, and optical scattering in subsequent spatial assessments. The expansion range can be predetermined based on the point spread characteristics of the depth sensor, monitoring distance, and calibration error. Point spread characteristics represent how light from a local location disperses to neighboring pixels during imaging. The range of influence under different conditions can be obtained through calibration, and the neighborhood size used for expansion can be set accordingly.

[0064] Specifically, morphological dilation can be used to incorporate pixels outside the boundary of the anomalous pixel region into the target pixel region according to a preset neighborhood. For example, if calibration results indicate that two layers of pixels around a saturated region may be affected by scattering, two layers of pixels can be expanded outward along the boundary of the anomalous pixel region. The expanded target pixel region includes the original anomalous pixel region and the newly added neighboring pixels. The inclusion of newly added pixels in the spatial judgment range does not mean that the depth data of all newly added pixels has been confirmed as invalid, but rather it is used to cover the unreliable measurement range that may exist near the anomalous boundary. The expansion size is determined based on the actual calibration results, and the number of example pixel layers does not constitute a fixed limit.

[0065] When determining the observation direction corresponding to the target pixel region, the angular coverage range corresponding to each pixel within the target pixel region can be obtained based on the imaging calibration parameters of the depth sensor, and the corresponding angular range can be transformed to a unified spatial coordinate system. Extending outward along the angular coverage range of each pixel, and then merging the corresponding spatial portions, the spatial range corresponding to the target pixel region is obtained. For irregular target pixel regions, mapping and merging can be performed separately according to the pixel set, preserving the region boundary features, rather than simply using a rectangular boundary covering the entire region for mapping.

[0066] The preset distance range can be defined by a minimum reliable distance and a maximum safe range. The minimum reliable distance corresponds to the near-end distance boundary that has been calibrated to meet monitoring requirements, while the maximum safe range corresponds to the far-end distance boundary that has been verified under the current acquisition conditions. By truncating the spatial range obtained by mapping the target pixel region according to the near-end and far-end distance boundaries, the observation spatial range used for subsequent intersection with the preset protected area can be obtained. Since invalid depth values ​​cannot provide reliable distances, the truncated range is not determined centered on invalid depth values, nor is it ignored because invalid depth values ​​show a greater distance.

[0067] For example, when a saturated pixel area forms near a metal roller, the abnormal pixel region composed of saturated pixels can be identified first. Then, the region can be expanded outwards based on the scattering influence range to obtain the target pixel region. Subsequently, spatial mapping is performed according to the observation direction of each pixel within the target pixel region, retaining the portion between the minimum confidence distance and the maximum safe range. The resulting observation spatial range covers not only the space corresponding to the original abnormal pixel but also the space near the abnormal boundary that may be affected, which is used to subsequently determine the unknown occupied area within the preset protected area.

[0068] Optionally, in the step of determining the unknown occupied area within the preset protection zone that cannot exclude the presence of a target based on the intersection relationship between the observation spatial range and the preset protection zone, the intersection area between the observation spatial range and the preset protection zone can also be determined; based on the spatial range of the intersection area and the preset minimum target size, it can be determined whether the intersection area can accommodate the preset minimum target; for intersection areas that cannot accommodate the preset minimum target, they are excluded from the unknown occupied area; based on the intersection areas that can accommodate the preset minimum target, the unknown occupied area is determined.

[0069] In this embodiment of the invention, the preset minimum target size refers to the lower limit of the target size that safety monitoring requirements can detect, and can be determined according to the human body parts or other targets that need to be protected. For example, for scenarios where hands approach dangerous parts, a corresponding minimum target size can be set according to hand detection requirements; for scenarios where personnel enter the robot's working area, a corresponding size can be set according to personnel detection requirements. The preset minimum target can be represented by a geometric model with corresponding dimensions, used to determine whether the space formed by local measurement failure is sufficient to completely accommodate the target that needs to be detected.

[0070] When determining intersecting areas, the spatial intersection of the observation area and the pre-defined protected area can be calculated, retaining the portion of space simultaneously located within both. For overlapping or connected intersecting portions, they can be merged for containment assessment, avoiding the division of a continuous area capable of jointly containing the target into multiple smaller areas for exclusion. For spatially separated intersecting portions, their shape and size can be analyzed separately.

[0071] When determining whether an intersecting region can accommodate a preset minimum target, the available space in different directions of the intersecting region can be compared with the size of the minimum target, and the shape of the region can be considered for judgment. Specifically, it can be checked whether there is a position within the intersecting region where the geometric model of the minimum target can be completely placed; when the target's pose affects the judgment result, it can also be checked whether there is a pose within the preset pose range that can be completely placed. When a position and pose that meet the conditions exist, it indicates that the minimum target may be completely located within an area that cannot be reliably measured, and the intersecting region is determined as an unknown occupied area.

[0072] The capacity assessment cannot be based solely on the total volume of the intersecting region or the number of invalid pixels. For example, a large, elongated region may not be able to accommodate a target with a certain lateral dimension, while a region with a small total volume but a concentrated shape may be sufficient to accommodate the target. Therefore, it is necessary to consider the continuous spatial range and boundary shape of the intersecting region. For cases where capacity cannot be clearly ruled out, they can be retained as unknown occupancy areas to avoid excluding areas that could potentially accommodate the target due to geometric approximation errors.

[0073] For intersecting regions that cannot accommodate a preset minimum target, they can be excluded from the unknown occupancy area if the surrounding effective observations meet the corresponding target detection requirements. The corresponding criterion is that a target reaching the preset minimum target size cannot be completely hidden within the intersecting region, and its excess portion can still be detected by surrounding effective observations. The exclusion operation only means that local data failure is not used as a sole basis for unknown occupancy; it does not change the invalidation determination of the original depth data, nor does it affect the occupancy determination of the actual target by the effective depth data.

[0074] For example, a continuous area of ​​invalid observation space forms near the black electrode and intersects with the preset protection zone on the feed side of the roller pressing equipment. If the intersecting area can completely accommodate the minimum target to be detected, then the intersecting area should be identified as an unknown occupied area, even if most of the surrounding pixels are valid. If the intersecting area is only a narrow local space that cannot accommodate the minimum target, and the surrounding valid observations can detect the portion of the target that exceeds it, then the intersecting area can be excluded. Thus, screening is based on whether the local invalid space can hide the target, avoiding judging its impact on the monitoring of the protection zone solely based on the overall proportion of invalid data.

[0075] Optionally, in the step of determining the occupancy status of the preset protected area based on the effective depth data and the unknown occupied area, the following steps can be taken: Based on the interference identification processing results, the target depth data that is being interfered with can be determined from the effective depth data, and the distance error range corresponding to the target depth data can be determined; based on the distance error range, the target location range corresponding to the target depth data can be expanded; it can be determined whether the expanded target location range intersects with the preset protected area to obtain an intersection judgment result; based on the intersection judgment result, the effective depth data other than the target depth data, and the unknown occupied area, the occupancy status of the preset protected area can be determined.

[0076] In this embodiment of the invention, depth data that still meets the validity conditions but whose measurement accuracy is affected by interference can be selected as target depth data based on the interference identification results. Specifically, data marked as valid in the aforementioned validity determination can be selected, and the distance error range can be determined according to the corresponding interference type and degree. The distance error range is used to limit the possible deviation of the actual target distance from the measured distance, so as to consider the possible location of the target in the occupancy determination.

[0077] The distance error range can be determined based on a pre-established calibration correspondence. During calibration, the target can be set at a known distance, depth data can be collected under different interference conditions, and the measured distance can be compared with the known distance to record the corresponding distance deviation. Based on the deviation distribution under different interference types, interference levels, and acquisition parameters, the corresponding near-end and far-end error limits are determined. In actual monitoring, based on the interference identification results corresponding to the target depth data and the acquisition parameters, the matching error limits are found to obtain the distance error range. The near-end and far-end error limits can be the same, or they can be set separately based on the calibration results.

[0078] When expanding the target location range based on the distance error range, the near-end error limit can be subtracted from the measured distance to obtain the near-end distance boundary of the possible target location, and the far-end error limit can be added to the measured distance to obtain the far-end distance boundary. Combining the observation direction and angular coverage of the corresponding pixels, the space between the two distance boundaries is determined as the expanded target location range. This process retains the original measured values ​​of the target depth data, incorporates distance deviations by expanding the location range involved in the judgment, and does not directly replace the measured values ​​with an estimated distance.

[0079] Subsequently, the expanded target location range can be compared with the spatial boundary of the preset protected area. If the two overlap or touch, the intersection judgment result is determined as intersection, and this is used as the basis for the existence of target occupation; if the two are completely separated, the intersection judgment result is determined as non-intersection. For multiple target depth data, the expanded target location range can be determined and judged separately. Any range that intersects with the preset protected area forms the corresponding basis for occupation.

[0080] For example, along the observation direction of a certain pixel, the preset protected area covers a distance range of 1 to 1.9 meters, while the measurement distance corresponding to the target depth data is 2 meters. If the calibration determines that both the near-end error limit and the far-end error limit are 0.2 meters, then the possible distance range of the target's location is 1.8 to 2.2 meters, which overlaps with the preset protected area's range of 1.8 to 1.9 meters. Therefore, even though the measurement location is outside the preset protected area, occupancy judgment still needs to be made based on the expanded range, assuming the presence of a target. The example values ​​are only for illustrating the processing method; the actual distance error range is determined by the corresponding calibration results.

[0081] When determining the overall occupancy status, the target depth data uses the expanded intersection judgment result, while the remaining effective depth data uses the corresponding target location judgment result. Unknown occupancy areas are still treated as having a target. If any of the above judgments indicates occupancy, the preset protected area is determined to be occupied; the expanded intersection result cannot be negated by the fact that the original measurement location of the target depth data is outside the protected area, or that the remaining effective depth data did not detect the target.

[0082] Optionally, in the step of outputting the corresponding industrial safety control signal based on the occupancy status, regional risk information can be determined based on the regional distribution of effective depth data and invalid depth data in multiple consecutive acquisition cycles. The regional risk information includes the effective coverage rate of the preset key area within the preset protection zone, the spatial range of the invalid connected area, and the duration of interference. Based on the interference type, interference level, and regional risk information, the industrial safety control level is determined, and the corresponding acquisition parameters and protection parameters are determined based on the industrial safety control level. The acquisition parameters include the modulation frequency and exposure parameters, and the protection parameters include the effective monitoring range, the adjustment range of the preset protection zone, and the allowable operating speed of the equipment. When the occupancy status indicates that the preset protection zone is occupied, or when the regional risk information meets the preset stop conditions, a safety stop signal is output. When the occupancy status indicates that the preset protection zone is idle and the regional risk information does not meet the preset stop conditions, monitoring is performed according to the acquisition parameters and protection parameters, and the corresponding industrial safety control signal is output based on the re-determined occupancy status.

[0083] In this embodiment of the invention, regional risk information is used to reflect the location, spatial range, and duration of interference within the protected area. Pre-defined critical areas can be delineated based on the location of the hazard source, the range of equipment movement, and personnel approach paths; for example, the area in front of the feed inlet of a roller press or near the dangerous movement range of a robot. The effective coverage rate of the critical area represents the proportion of effective observations covering the critical area that meet the target detection requirements. This can be calculated based on a pre-established spatial correspondence between pixels and critical areas, statistically analyzing the coverage portion that provides effective observations, and comparing it with the overall range of the critical area that needs to be monitored. A lower effective coverage rate indicates a larger portion of the critical area where the target location cannot be reliably confirmed.

[0084] The spatial extent of invalid connected regions can be determined based on the aforementioned abnormal pixel regions and their spatial mapping results, and the position, size, and overlap with key regions of each region can be recorded. For multiple invalid connected regions, the region with the largest extent can be selected as the basis for risk assessment to avoid situations where the overall proportion of invalid data is low, but the local continuous failure range is sufficient to hide the target. The duration of interference can be statistically analyzed by the position and overlap relationship of abnormal regions within a continuous acquisition period. For example, the timing can start when interference first appears in the corresponding region, and the cumulative duration can be updated when the corresponding interference continues to appear in subsequent periods.

[0085] Industrial safety control levels are used to differentiate the required control intensity under different perceived reliability conditions, and can include normal monitoring, disturbance monitoring, partial degradation, and safe shutdown. When determining the industrial safety control level, the effective coverage of the critical area, the range of the maximum ineffective connectivity area, and the duration of the disturbance can be compared with preset level conditions, and the corresponding level can be selected based on the type and severity of the disturbance. When multiple conditions point to different levels, a level with more stringent protection requirements can be selected. When the same level of disturbance occurs in a critical area, more stringent control can be applied than when it occurs in a non-critical area; if the disturbance continues to escalate or remains unresolved for an extended period, the control level can also be increased.

[0086] For example, normal monitoring can be maintained when the interference level is low, key areas still meet coverage requirements, and no target occupancy has occurred; when local measurement fluctuations increase but have not yet affected the necessary monitoring capabilities, disturbed monitoring can be initiated, and verification data collection can be increased; when the reliability of far-end measurements decreases but the near-end still meets protection requirements, local degradation can be initiated, and the protected area and the allowable operating speed of the equipment can be adjusted in conjunction with the effective monitoring range; when key areas cannot be reliably monitored, or when invalid areas have led to the protected area being determined to be occupied, a safe shutdown should be initiated. Local degradation cannot be used to offset already identified target occupancy or unknown occupancy.

[0087] Acquisition parameters are used to adjust the conditions for acquiring subsequent depth data. Modulation frequencies can include the main frequency used for normal measurements and the verification frequency used for consistency checks. Exposure parameters can include exposure duration and combinations of different exposure durations. To address multi-machine crosstalk, modulation frequencies can be switched according to a preset frequency detection sequence, selecting an allowable frequency with less interference. To address saturation interference, exposure duration can be shortened. To address low-echo conditions, exposure duration can be increased within permissible limits. For scenarios where highly reflective and low-reflective materials coexist, alternating short and long exposures can be used to check saturated and low-echo areas separately.

[0088] The acquisition parameters can be selected from pre-calibrated and verified parameter combinations, each corresponding to an applicable measurement range and acquisition time limit. After parameter adjustment, interference identification and validity determination need to be performed again using the newly acquired depth data; it cannot be assumed that the measurement has returned to reliability simply because frequency switching or exposure adjustment has been completed. If data meeting monitoring requirements cannot be obtained within the preset switching time limit, or if the parameter combination used exceeds the allowable range, this can be used as a safe stopping condition.

[0089] Protection parameters are used to match the equipment's operating conditions with the current effective monitoring capabilities. The effective monitoring range represents the distance range that can meet the target detection requirements under current acquisition parameters and interference conditions; the adjustment range of the preset protection zone represents the boundary of the protection zone that needs to be expanded or reconfigured; and the allowable operating speed of the equipment represents the upper speed limit that the equipment is allowed to reach under current protection conditions. When determining protection parameters, a pre-validated combination of parameters can be selected based on the effective monitoring range, measurement error, control response time, and equipment braking distance, enabling the equipment to perform necessary protective actions before the target approaches a dangerous position.

[0090] For example, if interference at the remote measurement shortens the effective monitoring range from 6 meters to 3 meters, a speed limit parameter matching the 3-meter effective range can be adopted, and the protection zone can be reconfigured within the reliable monitoring range. If the measurement error increases, a corresponding margin can be added to the boundary of the protection zone. A shortened range cannot be directly used as a basis for deleting the original protection requirement; if reducing the equipment speed and adjusting the protection zone still cannot meet the necessary detection and stopping distance requirements, a safety stop signal should be output.

[0091] Preset stop conditions may include critical area effective coverage falling below a minimum allowable value, invalid connectivity areas reaching a preset risk range, interference duration exceeding the allowable time limit, and parameter switching timeout. For example, when critical area effective coverage falls below 99.5%, an assessment can be made as to whether local degradation is necessary, and when it falls below 98%, a safe stop can be triggered. Specific thresholds should be determined based on the target size, the distribution of invalid areas, and equipment stop requirements; the risk of localized, continuous failures should not be ignored simply because the average coverage is high.

[0092] When a preset protected area is determined to be occupied, a safety stop signal is output directly, and the operating permit is not maintained on the grounds of low area risk. When a preset protected area is vacant but the area risk information meets the preset stop conditions, a safety stop signal is also output to handle situations where no target has been detected but monitoring capabilities are insufficient.

[0093] When the preset protected area is idle and the area risk information does not meet the preset stop conditions, subsequent data can be acquired according to the selected acquisition parameters, and the monitoring range, protected area range, and allowable operating speed of the equipment can be adjusted based on the protection parameters. The occupancy determination after parameter adjustment is based on the newly acquired data and the adjusted protected area, while simultaneously re-checking the area risk information; if occupancy is re-determined or the stop conditions are met, a safety stop signal is output; if the operating conditions continue to be met, a control signal corresponding to the current industrial safety control level is output. During parameter switching and re-determination, existing industrial safety control restrictions are not relaxed based on the idle result before adjustment.

[0094] Optionally, the above method can also switch the current industrial safety control level to a higher industrial safety control level when the regional risk information meets the entry conditions corresponding to the higher industrial safety control level; when the regional risk information meets the exit conditions corresponding to the current industrial safety control level, continuous verification is performed based on the re-collected depth data, and the risk level required for the exit conditions is lower than the risk level when entering the current industrial safety control level; within a continuous preset number of verification cycles, if the effective coverage, cross-frequency consistency, and occupancy status of the key area all meet the preset recovery conditions, and the self-inspection results are normal and the occupancy status of the preset protection zones determined by the two processing channels is consistent, the current industrial safety control level is reduced or the operation permit is restored; during continuous verification, it is prohibited to use historical depth data before entering the current industrial safety control level to determine that the preset protection zone is in an idle state; when the interference duration exceeds the preset time limit or the number of recovery failures reaches the preset number, the fault latching state is entered; after the preset manual reset conditions are met, the fault latching is released and continuous verification is re-executed, and a safety stop signal is maintained until the preset recovery conditions are met.

[0095] In this embodiment of the invention, the entry and exit conditions for industrial safety control levels can be set separately, enabling stricter controls to be implemented promptly when risks increase, and to be restored only after verification when risks decrease. Higher industrial safety control levels indicate levels requiring more stringent protection measures. When the effective coverage of a critical area decreases, the ineffective connectivity area expands, or the duration of interference increases, and the entry conditions for a higher level are met, the system can directly switch to the corresponding level without having to go through each intermediate level sequentially. For example, if the current system is at the disturbed monitoring level, but the effective coverage of the critical area has decreased to the safe stop condition, the system can directly enter the safe stop level.

[0096] Exit conditions are used to determine whether the basis for resuming verification is met. These requirements are more stringent than those for entering the current level, creating a hysteresis range between entry and exit. For interference evaluation indicators where higher values ​​indicate higher risk, the exit threshold can be set lower than the entry threshold. For effective coverage, the required coverage for exit can be set higher than the coverage that triggers degradation upon entry. When the risk is between these two thresholds, the current level is maintained to avoid frequent switching caused by interference fluctuating around the critical value. Meeting the exit conditions only indicates that continuous verification can begin; it does not directly remove the current control restrictions.

[0097] The verification cycle refers to the processing cycle required to complete a new set of data acquisition and recovery condition checks. It can include data acquisition under the main frequency, verification frequency, and corresponding exposure conditions. Within each verification cycle, the effective coverage of key areas is checked based on the reacquired data to ensure it meets recovery requirements, the measurement results at different modulation frequencies meet allowable deviations, and the preset protection zone is in an idle state. The exposure parameters used for acquisition should also be within the allowable range, and there should be no large-area saturation areas that affect the recovery judgment.

[0098] Self-testing can include internal reference channel checks, data acquisition and processing status checks, and safety output channel checks. For example, by comparing the measurement results of the internal reference target with the preset reference range, it can be determined whether the measurement channel is functioning correctly. The two processing channels can each determine the occupancy status of the preset protected area based on the current verification data, and then compare the results. The current verification cycle is only considered passed if the self-test is normal, both processing channels confirm the protected area is idle, and other recovery conditions are met simultaneously. If both processing channels determine the protected area is occupied, although the results are consistent, the idle conditions required for resumption of operation are not met.

[0099] Continuous verification can be achieved by recording the number of consecutive successful cycles. When all checks in each verification cycle pass, the number of successful cycles is accumulated. If any cycle experiences insufficient coverage, inconsistent cross-frequency measurements, protected area occupancy, self-inspection anomalies, or inconsistent results from two processing channels, the continuous verification is interrupted, and the number of consecutive successful cycles is re-accumulated. If the risk increases to a higher level of entry conditions during verification, the level upgrade is prioritized without waiting for the original verification process to complete.

[0100] For example, a control level reduction can be permitted only if the effective coverage rate of the key area reaches a preset recovery coverage rate, the measurement deviation between the main frequency and the calibration frequency is within the allowable range, the protected area remains idle, and both self-tests and dual-channel comparisons pass, all within 20 consecutive verification cycles. The recovery coverage rate can be set higher, for example, greater than 99.5%, depending on the entry conditions. The specific number of cycles and coverage rate are determined based on the acquisition cycle, interference variation characteristics, and protection requirements. After successful verification, the system can initially transition from the safe stop level to the disturbed monitoring level, and then resume normal monitoring based on subsequent monitoring results.

[0101] Continuous verification uses data re-acquired after entering the current industrial safety control level. Electronic equipment can distinguish between data before and after the switch by recording the level switch time or corresponding frame number, and clear or deactivate historical records previously used to prove the protected area was idle. Historical data can be retained for interference cause analysis and event tracing, but it cannot replace current valid observations, and should not be used to mask current measurement failures with idle results before the interference occurred.

[0102] The latch fault state is used to maintain stop control and prevent repeated automatic recovery processes when interference cannot be eliminated for a long time or when multiple recovery failures occur. The interference duration can be accumulated from the time the corresponding interference condition is met. The number of recovery failures can be counted based on recovery attempts; for example, a continuous verification is initiated after the exit condition is met, and a failure is recorded when the verification fails. The statistical time range and reset condition for the number of attempts can be preset. When the duration or number of failures reaches the corresponding limit, the latch fault state is entered, and a safety stop signal is continuously output. Even if a short period of normal measurement occurs subsequently, the latch will not be automatically released.

[0103] Preset manual reset conditions can include receiving a manual reset command, completing necessary maintenance, and ensuring the fault check meets the reset requirements. Once the manual reset conditions are met, only the fault latch is released and continuous verification is restarted; the operating permit is not immediately restored. During reverification, the safety stop signal remains active until the required number of consecutive passes is reached and all recovery conditions are met, at which point the corresponding operating permit is restored. This ensures that manual reset, measurement recovery confirmation, and operating permit restoration are executed separately, avoiding the direct release of protection by a single manual operation or momentary normal data.

[0104] Furthermore, such as Figure 2 As shown in the diagram, this embodiment of the invention also provides a flowchart of another industrial safety control method. The electronic device first loads verified modulation frequencies, exposure parameters, and corresponding usage ranges. It then collects depth data and optical measurement information under different modulation frequencies and exposure conditions. Through cross-frequency and cross-frame consistency analysis, it determines the type and degree of interference and generates corresponding safety validity labels. For data that can be corrected, the original observation labels are retained, and the correction results are used as estimated data. After meeting validity requirements, these are combined with the distance error range to participate in occupancy judgment. For data that cannot be reliably used, the corresponding depth values ​​are isolated, and the unknown occupancy area is determined based on the intersection relationship between the pixel's observation space range and the preset protected area. Subsequently, the occupancy status is determined by combining the effective depth data and the unknown occupancy area. Based on the effective coverage of key areas, invalid connected areas, and the duration of interference, the industrial safety control level is determined, and corresponding parameter adjustments or safety shutdown are executed. During the recovery process, continuous verification is performed using re-collected data. Only after successful verification is the control level lowered or the operating permit restored.

[0105] Furthermore, such as Figure 3As shown in the figure, this embodiment of the invention also provides a flowchart for determining the occupancy status of a preset protected area. In the figure, V0 represents reliable validity, V1 represents conditional validity, and both correspond to valid depth data. The data corresponding to V1 needs to be expanded to include the target location range before intersecting with the preset protected area. V2 represents invalid isolation; the corresponding depth value cannot be used as the basis for determining whether the protected area is vacant. Instead, it should be determined whether an unknown occupancy area has been formed based on the observation spatial range corresponding to the pixel. V3 represents unknown occupancy, corresponding to the aforementioned area where the presence of a target cannot be ruled out. The unknown occupancy entity in the figure is a spatial representation of an unknown occupancy area, used to participate in the protected area occupancy determination, and does not indicate that the shape of the actual target has been measured. The target occupancy formed by the valid depth data and the unknown occupancy jointly determine the occupancy status of the protected area. If either indicates occupancy, it is treated as occupancy. V4 indicates that there is a lack of sufficient reliable observations in the global or critical areas, making it impossible to complete a safety judgment, corresponding to situations where safety stopping conditions are met or more stringent protection measures are required.

[0106] Furthermore, such as Figure 4 As shown in the diagram, this embodiment of the invention also provides a schematic diagram of industrial safety control level switching. S0 to S4 in the diagram represent normal monitoring, disturbed monitoring, partial degradation, safe stop, and latching fault, respectively. As interference intensifies, effective coverage decreases, or unpredictable situations arise in critical areas, the industrial safety control level switches to a more stringent level; when the safe stop condition is met, it can directly enter S3. The range adjustment, protection zone adjustment, and speed limit control corresponding to S2 must meet the adjusted protection requirements and cannot replace the stop control when the protection zone is occupied. After the risk is reduced, the control recovery process is achieved through hysteresis conditions and continuous verification. Degradation is only allowed when coverage, cross-frequency consistency, occupancy status, self-test, and dual-channel comparison all meet the recovery requirements. After entering S4, manual reset conditions must be met to release the latch. After release, the system remains stopped and re-verified; the operating permit is not directly restored.

[0107] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0108] In one embodiment, an industrial safety control device is provided, which corresponds one-to-one with the industrial safety control methods described in the above embodiments. For example... Figure 5 As shown, the industrial safety control device includes an acquisition module 501, an identification module 502, a first determination module 503, a second determination module 504, and an output module 505. Detailed descriptions of each functional module are as follows: The acquisition module 501 is used to acquire depth data of the monitoring area, which includes a preset protected area; The identification module 502 is used to perform interference identification processing on the depth data, determine the valid depth data and invalid depth data in the depth data, and determine the observation space range of the pixels corresponding to the invalid depth data. The first determining module 503 is used to determine, based on the intersection relationship between the observation spatial range and the preset protection zone, an unknown occupied area within the preset protection zone where the presence of a target cannot be ruled out; The second determining module 504 is used to determine the occupancy status of the preset protected area based on the effective depth data and the unknown occupancy area, wherein the unknown occupancy area is determined to be occupied based on the presence of a target. The output module 505 is used to output a corresponding industrial safety control signal based on the occupancy status.

[0109] Optionally, the identification module 502 is further configured to: Obtain optical measurement information corresponding to the depth data, and determine interference features based on the optical measurement information and the depth data. The interference features include at least two of the following: echo intensity features, ambient light features, saturation features, cross-frequency consistency features, and cross-frame consistency features. Based on the interference characteristics, the interference type and degree of interference for the corresponding pixel are determined; Based on the type and degree of interference, the validity of the depth data is determined to identify the valid depth data and the invalid depth data.

[0110] Optionally, the identification module 502 is further configured to: Based on the spatial distribution of pixels corresponding to the invalid depth data, the abnormal pixel region is determined; According to a preset expansion rule, the abnormal pixel region is expanded to obtain the target pixel region; The spatial range corresponding to the target pixel region is determined based on the observation direction corresponding to the target pixel region; The portion of the spatial range that falls within a preset distance range is defined as the observation spatial range.

[0111] Optionally, the first determining module 503 is further configured to: Determine the area where the observation space range intersects with the preset protection zone; Based on the spatial range of the intersecting region and the preset minimum target size, determine whether the intersecting region can accommodate the preset minimum target; Intersecting regions that cannot accommodate the preset minimum target are excluded from the unknown occupied regions; The unknown occupied area is determined based on the intersecting area that can accommodate the preset minimum target.

[0112] Optionally, the second determining module 504 is further configured to: Based on the result of the interference identification process, the target depth data that has been interfered with is determined from the effective depth data, and the distance error range corresponding to the target depth data is determined. Based on the distance error range, the target location range corresponding to the target depth data is expanded; Determine whether the expanded target location range intersects with the preset protected area to obtain the intersection determination result; Based on the intersection judgment result, the effective depth data other than the target depth data, and the unknown occupied area, the occupancy status of the preset protected area is determined.

[0113] Optionally, the output module 505 is further configured to: Based on the effective depth data and the regional distribution of the invalid depth data in multiple consecutive acquisition cycles, regional risk information is determined. The regional risk information includes the effective coverage rate of the preset key area within the preset protection zone, the spatial range of the invalid connected area, and the duration of interference. Based on the type of interference, the degree of interference, and the regional risk information, the industrial safety control level is determined, and the corresponding acquisition parameters and protection parameters are determined based on the industrial safety control level. The acquisition parameters include modulation frequency and exposure parameters, and the protection parameters include effective monitoring range, adjustment range of the preset protection zone, and allowable operating speed of the equipment. When the occupancy status indicates that the preset protected area is occupied, or when the area risk information meets the preset stop conditions, a safety stop signal is output; When the occupancy status indicates that the preset protected area is vacant and the area risk information does not meet the preset stop condition, monitoring is performed according to the acquisition parameters and the protection parameters, and the corresponding industrial safety control signal is output according to the re-determined occupancy status.

[0114] Optionally, the device is further used for: When the risk information in the area meets the entry conditions corresponding to a higher industrial security control level, the current industrial security control level will be switched to the higher industrial security control level. When the risk information in the area meets the exit conditions corresponding to the current industrial safety control level, continuous verification is performed based on the re-collected depth data. The risk level required by the exit conditions is lower than the risk level when entering the current industrial safety control level. If, within a consecutive preset number of verification cycles, the effective coverage, cross-frequency consistency, and occupancy status of the key area all meet the preset recovery conditions, and the self-test results are normal and the occupancy status of the preset protected area determined by the two processing channels is consistent, the current industrial safety control level is reduced or the operation permit is restored. During the continuous verification, it is prohibited to use historical depth data prior to entering the current industrial safety control level to determine that the preset protected area is in an idle state; When the duration of the interference exceeds the preset time limit or the number of recovery failures reaches the preset number, the latching fault state is entered. After the preset manual reset conditions are met, the fault latch is released and the continuous verification is re-executed, and the safety stop signal is maintained until the preset recovery conditions are met.

[0115] Each module in the aforementioned industrial safety control device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0116] In one embodiment, a computer device is provided, which may be a terminal device, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, and network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a readable storage medium storing computer-readable instructions. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer-readable instructions implement an industrial safety control method. The readable storage medium provided in this embodiment includes both non-volatile and volatile readable storage media.

[0117] In one embodiment of the application, a readable storage medium is provided, which stores computer-readable instructions. When the computer-readable instructions are executed by a processor, they implement the steps of the industrial safety control method described above.

[0118] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware with computer-readable instructions. These computer-readable instructions can be stored in a non-volatile readable storage medium or a volatile readable storage medium. When executed, these computer-readable instructions can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0119] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0120] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. An industrial safety control method, characterized in that, The method includes: Acquire depth data of the monitoring area, which includes a preset protected area; The depth data is subjected to interference identification processing to determine the valid depth data and invalid depth data in the depth data, and the observation space range of the pixels corresponding to the invalid depth data is determined. Based on the intersection of the observation space range and the preset protection zone, unknown occupied areas within the preset protection zone that cannot be ruled out as having targets are determined; Based on the effective depth data and the unknown occupied area, the occupancy status of the preset protected area is determined, wherein the unknown occupied area is judged to be occupied based on the presence of a target; Based on the occupancy status, output the corresponding industrial safety control signal; The step of determining the unknown occupied area within the preset protection zone that cannot be ruled out as having a target, based on the intersection relationship between the observation spatial range and the preset protection zone, includes: Determine the area where the observation space range intersects with the preset protection zone; Based on the spatial range of the intersecting region and the preset minimum target size, determine whether the intersecting region can accommodate the preset minimum target; Intersecting regions that cannot accommodate the preset minimum target are excluded from the unknown occupied regions; The unknown occupied area is determined based on the intersecting area that can accommodate the preset minimum target.

2. The industrial safety control method according to claim 1, characterized in that, The interference identification processing of the depth data to determine the valid and invalid depth data includes: Obtain optical measurement information corresponding to the depth data, and determine interference features based on the optical measurement information and the depth data. The interference features include at least two of the following: echo intensity features, ambient light features, saturation features, cross-frequency consistency features, and cross-frame consistency features. Based on the interference characteristics, the interference type and degree of interference for the corresponding pixel are determined; Based on the type and degree of interference, the validity of the depth data is determined to identify the valid depth data and the invalid depth data.

3. The industrial safety control method according to claim 1, characterized in that, Determining the observation spatial range of the pixels corresponding to the invalid depth data includes: Based on the spatial distribution of pixels corresponding to the invalid depth data, the abnormal pixel region is determined; According to a preset expansion rule, the abnormal pixel region is expanded to obtain the target pixel region; The spatial range corresponding to the target pixel region is determined based on the observation direction corresponding to the target pixel region; The portion of the spatial range that falls within a preset distance range is defined as the observation spatial range.

4. The industrial safety control method according to claim 1, characterized in that, Determining the occupancy status of the preset protected area based on the effective depth data and the unknown occupied area includes: Based on the result of the interference identification process, the target depth data that has been interfered with is determined from the effective depth data, and the distance error range corresponding to the target depth data is determined. Based on the distance error range, the target location range corresponding to the target depth data is expanded; Determine whether the expanded target location range intersects with the preset protected area to obtain the intersection determination result; Based on the intersection judgment result, the effective depth data other than the target depth data, and the unknown occupied area, the occupancy status of the preset protected area is determined.

5. The industrial safety control method according to claim 2, characterized in that, The step of outputting a corresponding industrial safety control signal based on the occupancy status includes: Based on the effective depth data and the regional distribution of the invalid depth data in multiple consecutive acquisition cycles, regional risk information is determined. The regional risk information includes the effective coverage rate of the preset key area within the preset protection zone, the spatial range of the invalid connected area, and the duration of interference. Based on the type of interference, the degree of interference, and the regional risk information, the industrial safety control level is determined, and the corresponding acquisition parameters and protection parameters are determined based on the industrial safety control level. The acquisition parameters include modulation frequency and exposure parameters, and the protection parameters include effective monitoring range, adjustment range of the preset protection zone, and allowable operating speed of the equipment. When the occupancy status indicates that the preset protected area is occupied, or when the area risk information meets the preset stop conditions, a safety stop signal is output; When the occupancy status indicates that the preset protected area is vacant and the area risk information does not meet the preset stop condition, monitoring is performed according to the acquisition parameters and the protection parameters, and the corresponding industrial safety control signal is output according to the re-determined occupancy status.

6. The industrial safety control method according to claim 5, characterized in that, The method further includes: When the risk information in the area meets the entry conditions corresponding to a higher industrial security control level, the current industrial security control level will be switched to the higher industrial security control level. When the risk information in the area meets the exit conditions corresponding to the current industrial safety control level, continuous verification is performed based on the re-collected depth data. The risk level required by the exit conditions is lower than the risk level when entering the current industrial safety control level. If, within a consecutive preset number of verification cycles, the effective coverage, cross-frequency consistency, and occupancy status of the key area all meet the preset recovery conditions, and the self-test results are normal and the occupancy status of the preset protected area determined by the two processing channels is consistent, the current industrial safety control level is reduced or the operation permit is restored. During the continuous verification, it is prohibited to use historical depth data prior to entering the current industrial safety control level to determine that the preset protected area is in an idle state; When the duration of the interference exceeds the preset time limit or the number of recovery failures reaches the preset number, the latching fault state is entered. After the preset manual reset conditions are met, the fault latch is released and the continuous verification is re-executed, and the safety stop signal is maintained until the preset recovery conditions are met.

7. An industrial safety control device, characterized in that, The device includes: The acquisition module is used to acquire depth data of the monitoring area, which includes a preset protected area; The identification module is used to perform interference identification processing on the depth data, determine the valid depth data and invalid depth data in the depth data, and determine the observation space range of the pixels corresponding to the invalid depth data. The first determining module is used to determine, based on the intersection relationship between the observation spatial range and the preset protection zone, the unknown occupied area within the preset protection zone where the presence of a target cannot be ruled out; The second determining module is used to determine the occupancy status of the preset protected area based on the effective depth data and the unknown occupancy area, wherein the unknown occupancy area is determined to be occupied based on the presence of a target. The output module is used to output a corresponding industrial safety control signal based on the occupancy status; The first determining module is further configured to: Determine the area where the observation space range intersects with the preset protection zone; Based on the spatial range of the intersecting region and the preset minimum target size, determine whether the intersecting region can accommodate the preset minimum target; Intersecting regions that cannot accommodate the preset minimum target are excluded from the unknown occupied regions; The unknown occupied area is determined based on the intersecting area that can accommodate the preset minimum target.

8. A computer device comprising a memory, a processor, and computer-readable instructions stored in the memory and running on the processor, characterized in that, When the processor executes the computer-readable instructions, it implements the industrial safety control method as described in any one of claims 1 to 6.

9. A readable storage medium having computer-readable instructions stored thereon, characterized in that, When the computer-readable instructions are executed by a processor, they implement the industrial safety control method as described in any one of claims 1 to 6.

Images