Low-power consumption camera PIR trigger distance automatic testing device and method

Abstract

The application discloses a low-power consumption camera PIR trigger distance automatic testing device and method, comprising a test dark box, a light wall assembly, a measured low-power consumption camera positioning assembly, an infrared feedback collection assembly, an environment collection assembly, a device state collection assembly, a main controller, a memory and a report generation module. The main controller controls the far infrared heating pixel array to combine the equivalent moving heat source of the human body, vehicle or pet according to the real heat source characteristic library, pixel radiation calibration table and environment data, and corrects the output heat field through the infrared feedback closed loop. The test method adopts coarse scanning, boundary search and trigger probability fitting to determine the PIR trigger distance and response time, and combines the one-time test permission state record, pixel health state table and multi-residual anomaly identification record to eliminate abnormal data, so that the test authenticity, repeatability and traceability are improved.

Description

Technical Field

[0001] This invention relates to the fields of low-power camera testing, PIR sensor detection, and infrared heat source simulation, specifically to an automatic testing device and method for the PIR trigger distance of a low-power camera. Background Technology

[0002] Low-power surveillance cameras typically use PIR modules to detect targets with infrared radiation characteristics, such as humans, vehicles, and pets. Upon detecting a target entering the field of view, the camera is activated to acquire images, record video, or trigger an alarm. Because low-power cameras only retain a small portion of their detection circuitry in sleep mode, the trigger distance, trigger probability, and response time of the PIR module directly affect the actual monitoring range, missed detection rate, false trigger rate, and overall battery life of the device. Therefore, stable and repeatable testing of the PIR trigger distance of low-power cameras is necessary during product development, production testing, and factory inspection.

[0003] Current testing methods primarily rely on manual testing. Testers simulate human movement at different distances, observe whether the low-power camera is activated by the PIR module, and manually record the trigger status and response time. While this method requires less equipment investment, it is difficult to maintain consistency in the testers' movement speed, range of motion, clothing thickness, body temperature, direction of movement, and distance positioning. This results in significant differences in the trigger distance obtained for the same low-power camera under different testers or different test batches.

[0004] Some testing schemes use fixed targets, moving heat sources, constant-temperature lamps, or ordinary infrared lamp arrays to automatically test PIR modules. However, PIR modules primarily respond to changes in far-infrared thermal radiation, and ordinary infrared lamps, simple constant-temperature lamps, or single moving targets cannot realistically simulate the hotspot area, hotspot trajectory, radiation peak, rise time, and effective trigger duration formed when different heat sources such as humans, vehicles, and pets pass through the PIR field of view at different distances, moving speeds, and ambient temperatures. Especially for low-power cameras, the triggering characteristics of PIR modules in sleep / wake-up scenarios are also affected by factors such as background infrared radiation, ambient temperature, airflow disturbances, device sleep state, and wake-up acquisition channel delay. Simply using a fixed-power or fixed-distance heat source output method cannot guarantee the authenticity and traceability of test results.

[0005] The existing technology has the following shortcomings: 1. Manual testing suffers from poor consistency, with results heavily influenced by human factors. Existing manual testing requires testers to repeatedly simulate human movement at different distances. Because different testers vary in height, body temperature, clothing, walking speed, arm swing amplitude, movement route, and start and end times, even when testing the same low-power camera, different trigger distances and response times can easily be obtained, resulting in poor test repeatability.

[0006] 2. Existing automatic heat sources cannot accurately represent the real far-infrared thermal radiation process. Solutions using constant-temperature lamps, ordinary infrared lamps, or simple moving targets typically only provide a fixed heat source, fixed power, or simple sequential lighting effect. They cannot simultaneously simulate the hotspot area, hotspot height, peak thermal radiation, rise time, fall time, and hotspot center trajectory formed by real human bodies, vehicles, and pets at different distances and moving speeds, resulting in insufficient triggering equivalence for PIR modules.

[0007] 3. Lack of closed-loop calibration for individual differences and environmental drift in the lamp wall pixels. Far-infrared heating pixels vary in manufacturing error, aging degree, heating time, cooling time, and thermal inertia. If the lamp wall output is controlled solely based on a preset power meter without real-time acquisition of the actual radiation intensity of the array's output surface and closed-loop correction, the same control parameters may output different thermal fields at different pixels or under different ambient temperatures.

[0008] 4. The trigger distance determination method is crude and makes it difficult to obtain stable boundary indicators. Existing schemes mostly use "triggered" or "not triggered" as the conclusion of a single test, lacking repeated testing of candidate distances, trigger probability calculation, response time constraints, and probability curve fitting. For PIR responses near the critical distance, the results of a single trigger are random.

[0009] 5. The source of abnormal data is unclear, which can easily lead to misjudging equipment malfunctions or environmental disturbances as insufficient PIR performance. During the test, failure to trigger or false triggering may be caused by light wall radiation errors, abnormal hot spot trajectories, background infrared disturbances, airflow disturbances, the tested low-power camera not entering sleep mode, failure to release PIR signals, or asynchronous wake-up acquisition channels. Summary of the Invention

[0010] To address the problems existing in the prior art, this invention provides an automatic testing device and method for the PIR trigger distance of low-power cameras. It can perform equivalent synthesis of far-infrared thermal radiation based on the characteristics of real heat sources, and can combine infrared feedback closed-loop calibration, trigger distance adaptive search, test state locking, pixel health management and abnormal source isolation and recovery to complete automatic, standardized and traceable testing of the PIR trigger distance of low-power cameras.

[0011] To solve the above-mentioned technical problems, the present invention provides the following technical solution.

[0012] An automatic testing device for the PIR trigger distance of a low-power camera includes a test dark box, a light wall assembly, a positioning assembly for the low-power camera under test, an infrared feedback acquisition assembly, an environmental acquisition assembly, an equipment status acquisition assembly, a main controller, a memory, and a report generation module. The test chamber is a box-shaped structure with a closed test cavity. The inner wall of the test chamber is provided with a low-reflection, low-thermal-inertia lining layer to reduce the impact of external environmental thermal disturbance and thermal hysteresis of the inner wall of the test chamber on PIR triggering test. The first end of the test cavity is located on the side where the lamp wall assembly is located, and the second end of the test cavity is located on the side where the low-power camera positioning assembly under test is located. The lamp wall assembly is fixedly installed at the first end of the test cavity. The lamp wall assembly has an array emission surface facing the second end of the test cavity. The array emission surface is formed by M×N independently driveable far-infrared heating pixels arranged in a row and column matrix, where M and N are both integers greater than 1. The positioning component of the low-power camera under test is disposed at the second end of the test cavity. The positioning component of the low-power camera under test is used to fix the low-power camera under test and make the PIR window of the low-power camera under test face the array emission surface, and make the PIR detection center axis of the low-power camera under test pass through the geometric center of the array emission surface, or the lateral deviation and vertical deviation between the projection point of the PIR detection center axis on the array emission surface and the geometric center are not greater than half of the minimum center distance of adjacent far-infrared heating pixels and are not greater than 5mm. The infrared feedback acquisition component is installed on the top wall, side wall, or lateral unobstructed position of the lamp wall component of the test dark box. The detection field of view of the infrared feedback acquisition component covers the array emission surface and is used to acquire the actual radiation intensity and hot spot morphology of each far-infrared heating pixel. The environmental acquisition component is installed on the inner wall of the test chamber and located in the non-direct hot spot area between the lamp wall component and the positioning component of the low-power camera under test. It is used to collect environmental temperature, humidity, background infrared radiation and airflow data in the test chamber. The device status acquisition component is connected to or aligned with the status display area of ​​the low-power camera under test, and is used to acquire the PIR wake-up status and wake-up time of the low-power camera under test. The main controller is communicatively connected to the light wall assembly, infrared feedback acquisition assembly, environmental acquisition assembly, equipment status acquisition assembly, memory, and report generation module. The main controller is configured to: generate light wall control parameters based on the real heat source feature library and pixel radiation calibration data in the memory; control the far-infrared heating pixel array to synthesize an equivalent moving heat source; perform closed-loop correction of the driving parameters of the far-infrared heating pixels based on the actual radiation intensity acquired by the infrared feedback acquisition assembly; perform adaptive search for trigger distance based on the PIR wake-up status acquired by the equipment status acquisition assembly; and output the PIR trigger distance and response time test results by the report generation module.

[0013] Furthermore, the light wall assembly includes a plate-shaped support frame, a pixel mounting plate, a drive circuit board, and multiple far-infrared heating pixels; The plate-shaped support frame is fixed on the inner wall of the first end of the test dark box. The pixel mounting plate is a flat structure and is fixed on the side of the plate-shaped support frame near the positioning component of the low-power camera under test. Multiple far-infrared heating pixels are mounted on the pixel mounting plate with equal or preset unequal row and column spacing. Each far-infrared heating pixel is a sheet or block of ceramic far-infrared heating element, controllable blackbody radiation unit or micro heating film. The front of each far-infrared heating pixel constitutes part of the array emission surface, and the back of each far-infrared heating pixel is electrically connected to the drive circuit board. The driving circuit board has an independent driving channel corresponding to each far-infrared heating pixel. The independent driving channel can adjust the PWM duty cycle, driving current, lighting duration and cooling interval of the corresponding far-infrared heating pixel respectively. The low-power camera positioning component under test includes a rectangular base, a vertical adjustment bracket, and an angle adjustment seat. The rectangular base is fixed to the bottom wall of the test dark box, the vertical adjustment bracket is vertically set on the rectangular base, and the angle adjustment seat is installed on the side of the vertical adjustment bracket near the light wall component. The angle adjustment base has a mounting surface for supporting or clamping the low-power camera under test. The mounting surface is oriented towards the array emission surface. The angle adjustment base can adjust the pitch angle and / or horizontal deflection angle of the low-power camera under test, so that the center of the PIR window of different models of low-power cameras under test is aligned with the array center of the light wall assembly.

[0014] Furthermore, the infrared feedback acquisition component includes a thermal imager or a thermopile array; The infrared feedback acquisition component is mounted on the top or side wall of the test dark chamber by a fixed bracket. The detection axis of the infrared feedback acquisition component is inclined relative to the array emission surface of the lamp wall component and avoids the PIR detection center axis of the low-power camera under test, so as to acquire the radiation distribution of the array emission surface without obstructing the PIR detection field of view. The environmental acquisition component includes a temperature sensor, a humidity sensor, a background infrared radiation sensor, and an airflow sensor. The temperature sensor and humidity sensor are installed in the middle of the inner wall of the test chamber. The background infrared radiation sensor is positioned facing the non-heating area of ​​the inner wall of the test chamber. The airflow sensor is installed near the positioning component of the low-power camera under test and avoids the direct radiation area of ​​the light wall component. The device status acquisition component includes one or more of the following: serial port interface, USB interface, network log acquisition interface, GPIO acquisition interface, power current sampling unit and status indicator detection unit. The power current sampling unit is connected in series or coupled to the power supply line of the low-power camera under test, and the status indicator detection unit is located in front of or diagonally in front of the status indicator of the low-power camera under test. The memory stores a real heat source feature library, a pixel radiation calibration table, and a lamp wall control parameter database. The real heat source feature library stores the feature vector F of the real heat source when it passes through the PIR field of view according to the heat source category c, distance d, moving speed v, ambient temperature Ta, and moving direction θ. The feature vector F includes at least the target-background temperature difference ΔT, the equivalent hot spot area A, the hot spot center trajectory x(t), y(t), the thermal radiation peak value Rpeak, the rise time tr, the fall time tf, the PIR equivalent output peak value Vpeak, the effective trigger duration teff, and the noise baseline σ. The pixel radiation calibration table is established according to pixel number, driving parameters, ambient temperature, radiation intensity, heating time and cooling time; the lamp wall control parameter database establishes mapping records according to heat source type c, equivalent distance d, ambient temperature Ta, moving speed v, candidate pixel set Ω, duty cycle u, lighting duration τ and lighting order Q.

[0015] The equivalent distance referred to in this invention is the reference distance between the actual heat source and the PIR sensor, corresponding to the radiation intensity, hotspot area, hotspot distribution, and hotspot trajectory output by the light wall assembly, and not the actual physical distance between the light wall assembly and the low-power camera under test. The light wall control parameters corresponding to different equivalent distances are determined based on calibration data in the actual heat source feature library.

[0016] Furthermore, the main controller is configured to perform distance-power-trajectory matching and dynamic heat source trajectory generation; After setting the heat source category c and the equivalent distance d, the main controller reads the corresponding reference feature Fref from the real heat source feature library and generates a candidate pixel set Ω based on the geometric coordinates of each far-infrared heating pixel on the array emission surface. The main controller solves for the time-varying driving vector U={uij(t)} of each far-infrared heating pixel to minimize the error between the simulated hot spot feature Fsim synthesized by the lamp wall assembly and the reference feature Fref; where uij(t) represents the driving duty cycle or normalized driving intensity of the far-infrared heating pixel in the i-th row and j-th column at time t, Fsim includes at least the simulated radiation peak value Rsim, the simulated hot spot area Asim, the simulated rise time tr,sim, and the simulated PIR equivalent output peak value Vsim, and Fref includes at least the reference radiation peak value Rref, the reference hot spot area Asim, the reference rise time tr,ref, and the reference PIR equivalent output peak value Vref; The minimum error is evaluated using the objective function E. E=wR·|Rsim-Rref| / Rref+wA·|Asim-Aref| / Aref+wt·|tr,sim-tr,ref| / tr,ref+wV·|Vsim-Vref| / Vref+λ·[1 / (|T|·|GΩ|)]·Σ_{t∈T}Σ_{((i,j),(p,q))∈GΩ}[(uij(t)-upq(t))²], where T is the set of discrete sampling times of the driving vector U, GΩ is the set of unordered pairs of adjacent pixels formed by the four neighborhood relationships of top, bottom, left, and right within the candidate pixel set Ω, |T| and |GΩ| are the number of elements in the corresponding set, wR, wA, wt and wV are non-negative weights, λ is the power smoothing coefficient of adjacent pixels, and upq(t) is the driving intensity of the far-infrared heating pixel in the p-th row and q-th column paired with the far-infrared heating pixel in the i-th row and j-th column at time t; The main controller also generates a driving vector U based on the power smoothing constraint of adjacent pixels to avoid sudden changes between adjacent far-infrared heating pixels that do not conform to the actual heat source movement law. When simulating a lateral moving heat source, the main controller determines the switching period Δt=s / v of adjacent hot spot windows based on the target moving speed v and the spacing s of adjacent far-infrared heating pixels, and sequentially changes the set of candidate pixels Ω to be lit along the lateral direction of the array output surface. When simulating approaching or moving away from a heat source, the main controller achieves distance variation by changing the radiation power level and hot spot area corresponding to the equivalent distance d; when simulating a vehicle or pet heat source, the main controller forms an equivalent heat source pattern that is different from a human body heat source by changing the hot spot height, hot spot width, temperature distribution, and trajectory curvature. The dynamic heat source trajectory satisfies the following relationship: uij(t)=u0(t)·exp{-[(xij-xc(t))² / (2σx²)+(yij-yc(t))² / (2σy²)]}; where uij(t) is the driving intensity of the far-infrared heating pixel in the i-th row and j-th column at time t, u0(t) is the center driving intensity at time t, xij and yij are the array coordinates of the far-infrared heating pixel in the i-th row and j-th column, xc(t) and yc(t) are the center coordinates of the hot spot at time t, and σx and σy are the lateral diffusion parameter and longitudinal diffusion parameter related to the heat source type and equivalent distance, respectively.

[0017] Furthermore, the main controller is configured to perform closed-loop radiation control, baseline detection, and outlier data removal; During the output of the equivalent heat source by the lamp wall assembly, the infrared feedback acquisition assembly acquires the actual radiation intensity Lmeas of the array output surface. The main controller compares the actual radiation intensity Lmeas with the target radiation intensity Ltarget to obtain the error e(t) = Ltarget(t) - Lmeas(t), and corrects the driving parameters of the corresponding far-infrared heating pixel based on the error e(t). When the absolute value of the error e(t) is less than the preset error threshold ε within m consecutive sampling periods, the main controller determines that the current equivalent heat source has reached a stable state, where m is an integer from 3 to 10, and ε is 1% to 3% of the target radiation intensity Ltarget; when the absolute value of the error e(t) exceeds the preset error threshold ε, the main controller records the corresponding far-infrared heating pixel as a calibration abnormal pixel and recalibrates or removes the test data corresponding to the far-infrared heating pixel. Before each round of testing, the main controller monitors whether the low-power camera under test generates a PIR wake-up event through the device status acquisition component when the light wall component does not output a heat source. If a PIR wake-up event is generated, it is determined that there is background disturbance or the low-power camera under test is abnormal, and the test is stopped from proceeding to the valid test.

[0018] The present invention also provides an automatic testing method for the PIR trigger distance of a low-power camera, which is performed using the aforementioned automatic testing device for the PIR trigger distance of a low-power camera. The method includes the following steps: S1. Install the low-power camera under test on the positioning component of the low-power camera under test, so that the PIR window of the low-power camera under test faces the array emission surface of the light wall component, and align the PIR detection center axis with the geometric center of the array emission surface. S2. Read or build a real heat source feature library, which stores the feature vector F of a real heat source when it passes through the PIR field of view according to the heat source category c, distance d, moving speed v, ambient temperature Ta and moving direction θ. S3. Perform radiation calibration on each far-infrared heating pixel in the light wall assembly, establish a pixel radiation calibration table, and generate a light wall control parameter database based on the real heat source feature library and the pixel radiation calibration table. S4. Based on the target heat source type, target equivalent distance, target moving speed and current ambient temperature, read the corresponding candidate pixel set Ω, duty cycle u, lighting duration τ and lighting sequence Q from the light wall control parameter database, and control the light wall component to output the equivalent moving heat source. S5. Control or wait for the low-power camera under test to enter low-power sleep mode, and confirm through the device status acquisition component that the low-power camera under test only retains the wake-up function required for PIR detection. S6. At each candidate equivalent distance, control the lamp wall component to repeatedly output the equivalent moving heat source, and use the infrared feedback acquisition component to perform closed-loop correction on the actual radiation intensity of the lamp wall component. S7. Collect the PIR wake-up status and wake-up time of the low-power camera under test through the device status acquisition component, and calculate the trigger probability and response time under the candidate equivalent distance. S8. Perform adaptive search for trigger distance based on the trigger probability and response time under different candidate equivalent distances to determine the PIR trigger distance of the low-power camera under test, and generate a test report containing trigger distance, response time, repeatability statistics and abnormal records.

[0019] Furthermore, the adaptive search for triggering distance includes a coarse scan stage, a boundary search stage, and a probability fitting stage; During the coarse scan phase, equivalent moving heat sources are output sequentially according to a preset distance sequence from far to near: dmax, dmax-Δd, dmax-2Δd, ..., dmin. The trigger boundary interval [dnear, dfar] is determined based on whether the heat source is triggered. Here, dnear is the closer distance that meets the stable triggering condition, and dfar is the farther distance that is not triggered or has insufficient triggering probability. Furthermore, dnear is less than dfar. During the boundary search phase, a binary search or a staircase search is performed within the trigger boundary interval [dnear, dfar]. For each candidate equivalent distance d, the test is repeated k times, the number of triggers m is recorded, the trigger probability p(d) = m / k is calculated, and the mean response time μt and the standard deviation of response time σt are recorded. When p(d) ≥ Pth and μt ≤ τth, the candidate equivalent distance d is determined to meet the triggering requirements, where Pth is a preset trigger probability threshold and τth is a preset response time threshold. During the probability fitting stage, the trigger probability p(d) is fitted with a logistic curve, and the fitting relationship is p(d)=1 / [1+exp(a·(db))], where a is the trigger boundary slope parameter and b is the distance parameter near the trigger probability of 50%. When a is greater than 0, the distance index D(P0) with the trigger probability target value P0 is calculated according to D(P0)=b+ln(1 / P0-1) / a, where P0 is a preset probability target value between 0 and 1. The main controller outputs D50, D90 and / or D95 as PIR trigger distance indexes based on the fitted curve and generates confidence intervals. D50, D90 and D95 ​​correspond to D(0.50), D(0.90) and D(0.95) respectively. The response time is calculated as tresponse = twake - ton, where ton is the moment when the light wall component starts outputting the equivalent moving heat source, and twake is the moment when the low-power camera under test generates a PIR wake-up event. The ton and twake are recorded by the same master clock, or by the master controller sending a synchronization timestamp to the light wall component, the device status acquisition component, and the interface of the low-power camera under test.

[0020] Furthermore, before each test of the candidate equivalent distance d, the main controller first establishes a one-time test permission status record corresponding to the candidate equivalent distance d. The one-time test permission status record is used to characterize that the light wall component, the low-power camera under test, the test dark box environment, and the distance control parameters of this candidate equivalent distance test are all in the allowed test state. The one-time test permit status record is generated based on at least the light wall stability field, PIR silence field, environmental stability field, recovery release field, and distance parameter lock field. Wherein, the lamp wall stability field is set to allow when the absolute value of the error between the actual radiation intensity Lmeas and the target radiation intensity Ltarget is less than a preset error threshold ε within m consecutive sampling periods, where m is an integer from 3 to 10, and ε is 1% to 3% of the target radiation intensity; The PIR silence field is enabled when the GPIO acquisition interface does not detect a PIR wake-up transition, the power current sampling unit detects that the low-power camera under test is in the sleep current range, and the status indicator detection unit does not detect a change in the wake-up indication. The environmental stability field is enabled when the rate of change of ambient temperature in the test chamber is less than 0.2℃ / min, the rate of change of background infrared radiation is less than a preset background threshold, and the airflow speed is less than 0.2m / s; the recovery release field is enabled when the PIR wake-up signal after the last test returns to the silent state and continues for more than the recovery stabilization time Treset. The distance parameter locking field is set to allowed if it has not been modified after the candidate equivalent distance d, candidate cell set Ω, duty cycle u, lighting duration τ, and lighting order Q are written into the current test record; When the light wall stability field, PIR silence field, environmental stability field, recovery release field, and distance parameter lock field are all allowed, the main controller generates the one-time test permission status record and controls the light wall component to output an equivalent moving heat source; when any field is not allowed, the main controller does not output an equivalent moving heat source and records the reason for non-permission; after each equivalent moving heat source output ends, the one-time test permission status record becomes invalid, and a new one-time test permission status record is generated before the next test.

[0021] Furthermore, before the lamp wall component outputs an equivalent moving heat source, the main controller establishes a pixel health status table for far-infrared heating pixels and performs neighborhood replacement reconstruction for abnormal far-infrared heating pixels based on the pixel health status table. The pixel health status table stores the pixel number, row coordinates, column coordinates, radiation amplitude deviation, heating time deviation, cooling time deviation, and health status of each far-infrared heating pixel according to the row and column coordinates of the far-infrared heating pixels on the array emitting surface. When establishing the pixel health status table, the main controller applies low duty cycle, medium duty cycle, and high duty cycle drive signals to each far-infrared heating pixel, and the infrared feedback acquisition component acquires the actual radiation intensity Lij, heating time tup,ij, and cooling time tdown,ij of the far-infrared heating pixel under different duty cycles. The radiation amplitude deviation δLij is calculated according to δLij=|Lij-Lref,ij| / Lref,ij, where Lref,ij is the reference radiation intensity of the far-infrared heating pixel in the i-th row and j-th column of the pixel radiation calibration table; the heating time deviation δtup,ij is calculated according to δtup,ij=|tup,ij-tup,ref,ij| / tup,ref,ij, where tup,ref,ij is the reference heating time of the far-infrared heating pixel in the i-th row and j-th column of the pixel radiation calibration table; the cooling time deviation δtdown,ij is calculated according to δtdown,ij=|tdown,ij-tdown,ref,ij| / tdown,ref,ij, where tdown,ref,ij is the reference cooling time of the far-infrared heating pixel in the i-th row and j-th column of the pixel radiation calibration table; When δLij is greater than 10%, δtup,ij is greater than 15%, or δtdown,ij is greater than 15%, the main controller will mark the corresponding far-infrared heating pixel as an abnormal pixel. The health status includes one of the following: normal, insufficient radiation, heating lag, cooling lag, and failure. When the candidate pixel set Ω contains an abnormal pixel, the main controller selects at least two far-infrared heating pixels with normal health status from above, below, left, right and diagonally adjacent positions of the abnormal pixel as a replacement pixel set Ωr, and assigns compensation weights according to the distance between the replacement pixels and the abnormal pixels, so that the replacement pixel set Ωr jointly outputs the target radiation intensity originally corresponding to the abnormal pixel. After the neighborhood replacement reconstruction is completed, the main controller re-acquires the actual hot spot output by the lamp wall component through the infrared feedback acquisition component, and determines whether the compensated hot spot center deviation, hot spot area deviation, and cumulative thermal dose spectrum deviation are all less than the corresponding thresholds. The cumulative thermal dose spectrum deviation ΔHmap is calculated according to ΔHmap=Σ|Hmeas(i,j)-Href(i,j)| / ΣHref(i,j), where Hmeas(i,j) is the measured cumulative thermal dose of the far-infrared heating pixel in the i-th row and j-th column during one equivalent moving heat source output process, and Href(i,j) is the corresponding reference cumulative thermal dose. If all three deviations are less than the corresponding thresholds, the candidate equivalent distance test is allowed to continue. If any deviation exceeds the corresponding threshold, the candidate equivalent distance test is marked as lamp wall reconstruction failure data and is not included in the calculation of the trigger probability p(d).

[0022] Furthermore, during each test of the candidate equivalent distance d, the main controller synchronously calculates the multi-disability anomaly identification record and determines the source of the anomaly and the recovery action based on the multi-disability anomaly identification record; The multi-residual abnormality identification record includes at least the lamp wall radiation residual R1, cumulative thermal dose residual R2, ambient temperature residual R3, background infrared residual R4, dormant current residual R5, and wake-up time synchronization residual R6. Where R1 = |Ltarget - Lmeas| / Ltarget, Ltarget is the target radiation intensity, and Lmeas is the actual radiation intensity; R2 = |Hmeas - Href| / Href, Hmeas is the measured cumulative heat dose collected by the infrared feedback acquisition component during one equivalent moving heat source output process, and Href is the reference cumulative heat dose in the real heat source feature library under the corresponding heat source type, equivalent distance, and moving speed; R3 = |Ta - Ta,ref|, Ta ​​is the current ambient temperature in the test dark box, and Ta,ref is the reference ambient temperature in the real heat source feature library corresponding to the test conditions; R4 = |BIR - BIR,0| / BIR,0, where BIR is the current background infrared radiation and BIR,0 is the baseline background infrared radiation when the light wall component does not output an equivalent moving heat source; R5 = |Iidle - Iidle,ref| / Iidle,ref, where Iidle is the current sleep current of the low-power camera under test and Iidle,ref is the reference sleep current of the low-power camera under test; R6 = |twake,gpio - twake,current|, where twake,gpio is the PIR wake-up time recorded by the GPIO acquisition interface and twake,current is the current surge wake-up time recorded by the power supply current sampling unit. When R1 exceeds the limit but R3, R4, and R5 do not, the main controller determines that the source of the abnormality is an abnormal radiation output of the lamp wall component and re-executes the radiation calibration of the corresponding far-infrared heating pixel; when R2 exceeds the limit but R1 does not, the main controller determines that the source of the abnormality is an abnormality of the hot spot trajectory or the cumulative thermal process and recalculates the hot spot center trajectory, lighting sequence Q, and lighting duration τ. When R4 exceeds the limit, the main controller determines that there is background infrared disturbance in the test dark box and suspends the test until the background infrared radiation recovers to the baseline allowable range before re-performing baseline detection; when R5 exceeds the limit, the main controller determines that the low-power camera under test has not entered a valid low-power sleep state and resends the sleep command or outputs a device sleep abnormality prompt; when R6 exceeds the limit, the main controller determines that the device status acquisition component has a wake-up time acquisition synchronization error and resends the synchronization timestamp to the light wall component, the device status acquisition component, and the interface of the low-power camera under test. After the above recovery action, the main controller re-executes the test of the corresponding candidate equivalent distance d; if the same source of anomaly occurs consecutively a preset number of times, the data corresponding to the candidate equivalent distance d is marked as abnormal invalid data, and the source of anomaly, residual value and recovery action are recorded in the test report.

[0023] Beneficial effects: 1. Improve the realism of heat source simulation. This invention uses a light wall component composed of multiple far-infrared heating pixels and combines it with a real heat source feature library. Based on the heat source type, equivalent distance, moving speed, ambient temperature, and direction of movement, it generates corresponding candidate pixel sets, duty cycles, lighting durations, and lighting sequences. This enables a more realistic simulation of the far-infrared hot spot changes when different heat sources such as humans, vehicles, and pets pass through the PIR field of view.

[0024] 2. Improve the accuracy and repeatability of PIR trigger distance testing. This invention uses an infrared feedback acquisition component to collect the actual radiation intensity and hot spot morphology of the light wall component in real time, and the main controller performs closed-loop correction on the driving parameters of the far-infrared heating pixels, which can effectively eliminate output deviations caused by individual pixel differences, thermal inertia differences, and ambient temperature drift.

[0025] 3. It can objectively determine the critical trigger distance. This invention adopts an adaptive trigger distance search method that combines coarse scanning, boundary search, and probability fitting. It performs repeated tests multiple times at each candidate equivalent distance and makes a judgment based on the trigger probability and response time, finally outputting trigger distance indicators of D50, D90, and / or D95.

[0026] 4. Reduce the risk of invalid test data contamination. This invention generates a one-time test permission status record before each candidate equivalent distance test. The equivalent moving heat source is only allowed to be output when the conditions of lamp wall stability, PIR silence, environmental stability, recovery release, and distance parameter locking are all met.

[0027] 5. Improve long-term reliability and anomaly traceability. This invention identifies far-infrared heating pixels with insufficient radiation, delayed heating, delayed cooling, or failure through a pixel health status table, and maintains the continuity of hot spot output through neighborhood substitution reconstruction; at the same time, it identifies the source of anomalies through multi-residual anomaly identification records and performs corresponding recovery actions. Attached Figure Description

[0028] Figure 1 This is a system block diagram of the low-power camera PIR trigger distance automatic testing device of the present invention; Figure 2 This is a schematic diagram of the equivalent moving heat source trajectory of the present invention; Figure 3 This is a flowchart of the adaptive search process for triggering distance in this invention; Figure 4 This is a flowchart of the test license lockout, pixel health status table, and multi-disability abnormality identification and isolation process of the present invention. Detailed Implementation

[0029] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are intended to illustrate possible implementations of the invention and not to limit the scope of protection of the invention; where there is no conflict, the technical features in the following embodiments can be combined with each other.

[0030] Example 1 This embodiment uses a battery-powered low-power surveillance camera as the test camera to measure its PIR module's maximum effective trigger distance, trigger probability, and response time in a scenario involving lateral human movement. This embodiment employs the aforementioned automatic testing device for the low-power camera's PIR trigger distance.

[0031] The test chamber 1 is a rectangular box structure, forming a closed test cavity 11 inside. The inner wall of the test chamber 1 is covered with a black low-reflection, low-thermal-inertia lining layer 12. The lining layer 12 uses a black matte coating, a low-reflection velvet panel, or a composite heat insulation board with a matte finish to reduce the influence of external visible light, infrared background radiation, and thermal hysteresis of the inner wall of the chamber on the PIR test results. The side where the lamp wall assembly 2 is located is designated as the first end of the test cavity 11, and the side where the low-power camera positioning assembly 3 under test is located is designated as the second end of the test cavity 11. The lamp wall assembly 2 is fixedly installed on the inner wall of the first end of the test chamber 1, and the low-power camera positioning assembly 3 under test is fixedly installed on the bottom wall of the second end of the test chamber 1.

[0032] The light wall assembly 2 includes a plate-shaped support frame 21, a pixel mounting plate 22, a drive circuit board, and 16×16 far-infrared heating pixels. The plate-shaped support frame 21 is a rectangular frame structure, fixed to the inner wall of the first end of the test dark chamber 1. The pixel mounting plate 22 is a flat plate structure, fixed to the side of the plate-shaped support frame 21 near the low-power camera positioning assembly 3 under test. 256 far-infrared heating pixels are arranged in a 16x16 matrix on the pixel mounting plate 22, with a center-to-center distance of 20mm between adjacent far-infrared heating pixels. Each far-infrared heating pixel is a sheet-shaped ceramic far-infrared heating element, with its front facing the low-power camera positioning assembly 3 under test. The front faces of all far-infrared heating pixels together form an array emission surface. The drive circuit board is located on the side of the pixel mounting plate 22 away from the array emission surface and is electrically connected to each far-infrared heating pixel via wires or solder pads.

[0033] Each far-infrared heating pixel has an independent drive channel, which can adjust the PWM duty cycle, drive current, illumination duration, and cooling interval of that far-infrared heating pixel. The PWM duty cycle can be adjusted from 0% to 100%, the drive current from 0mA to 300mA, and the single illumination duration is determined based on the target's moving speed, the center distance between adjacent far-infrared heating pixels, and the dynamic response characteristics of the far-infrared heating pixels. The cooling interval is from 100ms to 5000ms. Each far-infrared heating pixel is used to output an equivalent hot spot corresponding to the far-infrared thermal radiation characteristics of the human body.

[0034] The positioning assembly 3 for the low-power camera under test includes a rectangular base 31, a vertical adjustment bracket 32, and an angle adjustment seat 33. The rectangular base 31 is fixedly installed on the bottom wall of the second end of the test dark box 1; the vertical adjustment bracket 32 ​​is vertically mounted on the rectangular base 31; and the angle adjustment seat 33 is installed on the side of the vertical adjustment bracket 32 ​​near the lamp wall assembly 2. After the low-power camera under test 4 is fixed, its PIR window 41 is located on the side of the angle adjustment seat 33 near the lamp wall assembly 2 and faces the array emission surface of the lamp wall assembly 2. The height of the low-power camera under test 4 is adjusted by the vertical adjustment bracket 32, and the pitch angle and horizontal deflection angle of the low-power camera under test 4 are adjusted by the angle adjustment seat 33, so that the center of the PIR window 41 is basically aligned with the geometric center of the array emission surface of the lamp wall assembly 2. The height deviation between the center of the PIR window 41 and the geometric center of the array emission surface is controlled within ±5mm, the horizontal deviation is controlled within ±5mm, and the pitch angle deviation is controlled within ±2°.

[0035] The infrared feedback acquisition component 5 employs a thermal imager. The thermal imager is mounted on the inner top wall of the test chamber 1 via a fixed bracket, located in the area above the lamp wall assembly 2 and the low-power camera under test positioning component 3. The detection axis of the thermal imager is tilted relative to the array emission surface of the lamp wall assembly 2 at an angle of 20° to 45°, and the thermal imager is not located on the PIR detection center axis of the low-power camera under test 4, thus avoiding obstruction of the PIR detection field of view. The detection field of view of the thermal imager covers the entire array emission surface of the lamp wall assembly 2, used to acquire in real time the actual radiation intensity, hot spot area, and hot spot center position of each far-infrared heating pixel.

[0036] The environmental acquisition component 6 includes a temperature sensor, a humidity sensor, a background infrared radiation sensor, and an airflow sensor. The temperature and humidity sensors are installed in the middle of one side of the inner wall of the test chamber 1, at a height that is basically the same as the height of the PIR window 41 of the low-power camera 4 under test. The background infrared radiation sensor is installed on the other side of the inner wall of the test chamber 1, facing the non-heat-generating area of ​​the inner wall of the test chamber 1. The airflow sensor is installed near the positioning component 3 of the low-power camera under test, avoiding the direct radiation area of ​​the light wall component 2.

[0037] The device status acquisition component 7 includes a GPIO acquisition interface, a power supply current sampling unit, and a status indicator detection unit. The GPIO acquisition interface is connected to the PIR wake-up signal test terminal of the low-power camera under test 4; the power supply current sampling unit is connected in series in the power supply line of the low-power camera under test 4; the status indicator detection unit is a photoelectric sensor, located diagonally in front of the status indicator of the low-power camera under test 4. The above three detection methods can be used individually or simultaneously; when used simultaneously, the wake-up time detected by the GPIO acquisition interface is taken as the priority wake-up time, and the power supply current sampling unit and the status indicator detection unit serve as verification channels.

[0038] The real thermal source feature library 91 can be established through pre-measured data or imported from existing calibration data. Taking the lateral movement of a human body as an example, multiple calibration distances are set within a range of 3m to 15m, multiple movement speeds are set within a speed range of 0.3m / s to 1.2m / s, and multiple environmental points are set within an ambient temperature range of 20℃ to 35℃. A calibrated thermal imager is used to collect the hot spot area, hot spot center trajectory, radiation peak value, rise time, fall time, and effective trigger duration when the human body passes through the PIR field of view. At the same time, a pre-calibrated standard reference PIR prototype is used to record the corresponding normalized PIR response peak value Vpeak. The standard reference PIR prototype is set separately from the low-power camera 4 under test and does not participate in the subsequent trigger determination of the device under test. The above-mentioned collected data is normalized and stored as reference feature Fref.

[0039] The equivalent distance referred to in this invention is the reference distance between the actual heat source and the PIR sensor, corresponding to the radiation intensity, hotspot area, hotspot distribution, and hotspot trajectory output by the light wall assembly, and not the actual physical distance between the light wall assembly and the low-power camera under test. The light wall control parameters corresponding to different equivalent distances are determined based on calibration data in the actual heat source feature library.

[0040] The main controller establishes a mapping relationship between thermal radiation characteristics and PIR response peak value based on a pre-calibrated standard reference PIR prototype, and calculates the simulated PIR equivalent output peak value Vsim based on the simulated radiation peak value, hot spot area, rise time and effective trigger duration formed by the lamp wall assembly.

[0041] Before the test begins, the main controller 8 performs device initialization. The operator fixes the low-power camera 4 under test on the angle adjustment seat 33 and adjusts the angle adjustment seat 33 so that the PIR window 41 faces the array emission surface of the light wall assembly 2. Then, the door of the test dark box 1 is closed, making the test chamber 11 a closed space. The main controller 8 reads data from the temperature sensor, humidity sensor, background infrared radiation sensor, and airflow sensor. When the ambient temperature inside the test chamber 11 is stable at 25℃±2℃, the relative humidity is stable between 40% and 70%, the background infrared radiation fluctuation is less than the preset background threshold, and the airflow speed is less than 0.2m / s, the light wall calibration step begins.

[0042] During the wall light calibration, the main controller 8 performs radiation intensity calibration on each far-infrared heating pixel individually. Taking the far-infrared heating pixel in the i-th row and j-th column as an example, the main controller 8 sequentially sets the PWM duty cycle u of the far-infrared heating pixel to 10%, 20%, 40%, 60%, 80%, and 100%, maintaining each duty cycle for 2 seconds. The infrared feedback acquisition component 5 records the actual radiation intensity Lij of the far-infrared heating pixel under different duty cycles. Based on the acquisition results, the main controller 8 establishes a radiation model for the far-infrared heating pixel: Lij(u, Ta) = aij(Ta)·u² + bij(Ta)·u + cij(Ta). Wherein, Lij(u, Ta) represents the actual radiation intensity of the far-infrared heating pixel in the i-th row and j-th column under the ambient temperature Ta and duty cycle u; aij(Ta), bij(Ta), and cij(Ta) represent the quadratic coefficient, linear coefficient, and constant coefficient related to the ambient temperature Ta, respectively; and u represents the PWM duty cycle of the far-infrared heating pixel.

[0043] The human heat source category is set as an adult moving laterally, the target moving speed is set to 0.5 m / s, and the initial ambient temperature is set to 25℃. The main controller 8 reads the reference feature Fref of the human heat source under the corresponding ambient temperature and moving speed from the real heat source feature library 91, and generates a candidate pixel set Ω based on the geometric coordinates of each far-infrared heating pixel in the light wall component 2. For the human lateral movement scenario, the candidate pixel set Ω consists of several rows of pixels in the upper middle part of the array output surface and their adjacent pixels, used to form an equivalent hot spot window with a height and width close to the human hot spot projection.

[0044] During the dynamic human body heat source generation process, the main controller 8 controls the illumination intensity of the far-infrared heating pixels according to the two-dimensional continuous hot spot model: uij(t)=u0(t)·exp{-[(xij-xc(t))² / (2σx²)+(yij-yc(t))² / (2σy²)]}. Where uij(t) is the driving intensity of the far-infrared heating pixel in the i-th row and j-th column at time t; u0(t) is the center driving intensity at time t; xij and yij are the coordinates of the far-infrared heating pixel in the i-th row and j-th column on the array's output surface; xc(t) and yc(t) are the center coordinates of the hot spot at time t; σx is the lateral diffusion parameter of the hot spot; and σy is the longitudinal diffusion parameter of the hot spot. The center distance s between adjacent far-infrared heating pixels is 20mm, and the target moving speed v is 0.5m / s. Therefore, the switching period Δt between adjacent hot spot windows is Δt=s / v=0.04s, or 40ms.

[0045] In this embodiment, the main controller 8 forms a discrete sampling time set T with a sampling period of 20ms, and forms a set of unordered adjacent pixel pairs GΩ using the four-neighbor relationship of top, bottom, left, and right. The preferred weights of the objective function E are set as wR=0.30, wA=0.20, wt=0.20, wV=0.20, and λ=0.10. These weights are used to achieve a balance between radiation peak value, hot spot area, rise time, reference PIR response peak value, and adjacent pixel smoothing. When the test object is a vehicle or pet heat source, the total value of wR, wA, wt, and wV can be kept at 0.90, and the corresponding weights can be adjusted within ±0.10 according to the importance of hot spot area or rise time.

[0046] The main controller 8 controls the low-power camera under test 4 to enter a low-power sleep mode, and confirms through the power current sampling unit that the operating current of the low-power camera under test 4 has decreased to the sleep current range, for example, less than 1mA; at the same time, it confirms through the GPIO acquisition interface that the PIR wake-up signal is in a silent state. Before entering the effective test, the light wall component 2 does not output an equivalent heat source, and the main controller 8 continuously monitors the PIR wake-up status of the low-power camera under test 4 for 30 seconds; if there is no wake-up event, the baseline is determined to be normal.

[0047] The trigger distance test is performed using a three-step process: coarse scanning, boundary search, and probability fitting. During the coarse scanning phase, the main controller 8 controls the light wall assembly 2 to output the equivalent heat source for lateral human movement according to an equivalent distance sequence from far to near. The preset distance sequence is 12m, 11m, 10m, 9m, 8m, 7m, 6m, 5m, 4m, and 3m. Each equivalent distance is tested three times. Before each test, the tested low-power camera 4 is allowed to return to a sleep / silent state, and a recovery stabilization time Treset is allowed. In this embodiment, Treset is 10s. The trigger boundary interval [dnear, dfar] is determined through coarse scanning.

[0048] During the boundary search phase, the main controller 8 performs a binary search or a staircase search between [dnear, dfar]. Each candidate equivalent distance d is tested 10 times. During each test, the main controller 8 records the moment ton when the light wall component 2 begins to output the equivalent moving heat source, and records the moment twake when the tested low-power camera 4 generates a PIR wake-up event through the device status acquisition component 7. The response time is calculated as tresponse = twake - ton. For each candidate equivalent distance d, the main controller 8 records the number of triggers m in the 10 tests and calculates the trigger probability p(d) = m / 10. When p(d) ≥ 90% and the average response time μt ≤ 3s, the candidate equivalent distance d is deemed to meet the triggering requirements.

[0049] During the output of the equivalent moving heat source by the light wall component 2, the infrared feedback acquisition component 5 detects the actual radiation intensity Lmeas of the array's emitting surface in real time and sends the detection result to the main controller 8. The main controller 8 compares the actual radiation intensity Lmeas with the target radiation intensity Ltarget to obtain the error e(t) = Ltarget(t) - Lmeas(t), and corrects the duty cycle or driving current of the corresponding far-infrared heating pixel based on the error e(t). When |e(t)| is less than 2% of the target radiation intensity for 5 consecutive sampling periods, the equivalent heat source output is considered stable; if |e(t)| continuously exceeds 3% of the target radiation intensity, the main controller 8 marks this test as an abnormal light wall output, and the test data is not included in the trigger probability calculation.

[0050] After the boundary search is completed, the main controller 8 performs logistic curve fitting on the trigger probability p(d) under different candidate equivalent distances: p(d) = 1 / [1 + exp(a·(db))]. Here, a is the trigger boundary slope parameter, b is the distance parameter when the trigger probability is approximately 50%, and d is the equivalent distance. Based on the fitted curve, the main controller 8 outputs D50, D90, and D95 ​​as the PIR trigger distance indicators of the tested low-power camera 4, where D50, D90, and D95 ​​correspond to D(0.50), D(0.90), and D(0.95), respectively, and the report generation module 10 generates a test report.

[0051] Example 2 This embodiment uses the same low-power camera PIR trigger distance automatic testing device as Embodiment 1; however, this embodiment is used to simulate vehicle heat source and pet heat source, and introduces a one-time test permission status record before each candidate equivalent distance test. When the light wall component 2 is used for a long time or there are abnormal far-infrared heating pixels, a pixel health status table and neighborhood replacement reconstruction are introduced. During the test, the abnormal identification record of multiple residuals is used to distinguish the abnormal sources of light wall, hot spot trajectory, background infrared disturbance, the tested low-power camera not sleeping, and asynchronous wake-up acquisition.

[0052] In this embodiment, the light wall assembly 2 includes a plate-shaped support frame 21, a pixel mounting plate 22, a drive circuit board, and 20×20 far-infrared heating pixels. The far-infrared heating pixels are arranged in a 20x20 matrix, with a center-to-center distance of 18mm between adjacent far-infrared heating pixels. Each far-infrared heating pixel is a sheet-shaped ceramic far-infrared heating element, with its front facing the low-power camera positioning assembly 3 under test, and its back electrically connected to the corresponding independent drive channel of the drive circuit board.

[0053] After the test begins, the main controller 8 reads the real heat source feature library 91 from the memory 9. The real heat source feature library 91 pre-stores the feature vectors of vehicle heat sources and pet heat sources passing through the PIR field of view under different equivalent distances, different moving speeds, different ambient temperatures, and different directions of movement. The vehicle heat source feature vector includes the temperature difference between the vehicle target and the background ΔTvehicle, the vehicle's equivalent hot spot area Avehicle, the vehicle's hot spot center trajectory xvehicle(t), yvehicle(t), the vehicle's peak thermal radiation Rpeak,vehicle, the vehicle's rise time tr,vehicle, the vehicle's fall time tf,vehicle, the vehicle's effective trigger duration teff,vehicle, and the vehicle's noise baseline σvehicle. The pet heat source feature vector includes the pet target and background temperature difference ΔTpet, the pet equivalent hot spot area Apet, the pet hot spot center trajectory xpet(t), ypet(t), the pet thermal radiation peak Rpeak,pet, the pet rise time tr,pet, the pet fall time tf,pet, the pet effective trigger duration teff,pet, and the pet noise baseline σpet.

[0054] The simulation parameters for the vehicle heat source are set as follows: target moving speed is 1.0 m / s, and the equivalent distance sequence is 15 m, 14 m, 13 m, 12 m, 11 m, 10 m, 9 m, 8 m, 7 m, and 6 m. The vehicle heat spot window is located in the central region of the array's emission surface. The heat spot height corresponds to 6 to 9 rows of far-infrared heating pixels, and the heat spot width corresponds to 8 to 12 columns of far-infrared heating pixels. The heat spot center trajectory is a horizontal straight line or a gently curved trajectory. The simulation parameters for the pet heat source are set as follows: target moving speed is 0.4 m / s, and the equivalent distance sequence is 10 m, 9 m, 8 m, 7 m, 6 m, 5 m, 4 m, 3 m, and 2 m. The pet heat spot window is located in the lower region of the array's emission surface. The heat spot height corresponds to 2 to 4 rows of far-infrared heating pixels, and the heat spot width corresponds to 3 to 5 columns of far-infrared heating pixels. The heat spot center trajectory is an irregular curved trajectory.

[0055] Before performing vehicle or pet heat source tests, the main controller 8 first establishes a pixel health status table. For the far-infrared heating pixel p(i,j) in the i-th row and j-th column, the main controller 8 sequentially applies drive signals with three duty cycles of 20%, 50%, and 80%, each duty cycle lasting 2 seconds. The infrared feedback acquisition component 5 acquires the actual radiation intensity Lij, heating time tup,ij, and cooling time tdown,ij of the far-infrared heating pixel at each duty cycle. The heating time tup,ij refers to the time from the start of driving the far-infrared heating pixel until the actual radiation intensity reaches 90% of the target radiation intensity; the cooling time tdown,ij refers to the time after driving stops when the actual radiation intensity drops to 10% of the target radiation intensity.

[0056] The main controller 8 calculates the radiation amplitude deviation δLij = |Lij - Lref,ij| / Lref,ij, the heating time deviation δtup,ij = |tup,ij - tup,ref,ij| / tup,ref,ij, and the cooling time deviation δtdown,ij = |tdown,ij - tdown,ref,ij| / tdown,ref,ij. When δLij ≤ 10%, δtup,ij ≤ 15%, and δtdown,ij ≤ 15%, the main controller 8 marks the corresponding far-infrared heating pixel as normal. When δLij > 10% and the actual radiation intensity is lower than the reference radiation intensity, it is marked as insufficient radiation. When δtup,ij > 15%, it is marked as heating hysteresis. When δtdown,ij > 15%, it is marked as cooling hysteresis. When the far-infrared heating pixel still cannot reach 50% of the reference radiation intensity at an 80% duty cycle, it is marked as failed.

[0057] When the candidate pixel set Ω contains an abnormal far-infrared heating pixel, the main controller 8 does not directly drive the abnormal far-infrared heating pixel. Instead, it selects at least two far-infrared heating pixels with normal health status from above, below, to the left, to the right, and four diagonally adjacent positions of the abnormal far-infrared heating pixel as a replacement pixel set Ωr. Preferably, the replacement pixel set Ωr includes the four normal far-infrared heating pixels closest to the abnormal far-infrared heating pixel. Let the target radiation intensity that the abnormal far-infrared heating pixel should originally output be Lbad,target, and the distance between the q-th replacement pixel in the replacement pixel set Ωr and the abnormal far-infrared heating pixel be rq. Then, the compensation weight of the q-th replacement pixel is wq=(1 / rq) / Σ(1 / rq), and the compensation target radiation intensity of the q-th replacement pixel is wq·Lbad,target.

[0058] After completing the neighborhood replacement reconstruction, the main controller 8 re-acquires the actual hot spot output by the lamp wall component 2 through the infrared feedback acquisition component 5 and performs three verifications: First, the hot spot center deviation is no greater than 0.5 times the center distance of one pixel; second, the hot spot area deviation is no greater than 5%; third, the cumulative thermal dose map deviation ΔHmap is no greater than 5%, where ΔHmap = Σ|Hmeas(i,j) - Href(i,j)| / ΣHref(i,j). If all three verifications are qualified, the main controller 8 allows the corresponding candidate equivalent distance test to continue. If any verification fails, the main controller 8 marks the candidate equivalent distance test as lamp wall reconstruction failure data and does not include it in the trigger probability calculation.

[0059] Before each candidate equivalent distance test, the main controller 8 establishes a one-time test permission status record corresponding to that candidate equivalent distance. This one-time test permission status record is generated by the main controller 8 based on the status of the light wall components, the status of the low-power camera under test, the status of the test darkroom environment, and the distance control parameters for this test. The one-time test permission status record includes at least the light wall stability field, the PIR silence field, the environmental stability field, the recovery release field, and the distance parameter lock field. The criterion for the light wall stability field is: within 5 consecutive sampling periods, the absolute value of the error between the actual radiation intensity Lmeas and the target radiation intensity Ltarget is less than 2% of the target radiation intensity.

[0060] The PIR silence field is judged as follows: the GPIO acquisition interface does not detect a PIR wake-up transition, the power current sampling unit detects that the tested low-power camera 4 is in the sleep current range, and the status indicator detection unit does not detect a change in the wake-up indication. The environmental stability field is judged as follows: the ambient temperature change rate in the test dark box 1 is less than 0.2℃ / min, the background infrared radiation change rate is less than the preset background threshold, and the airflow speed is less than 0.2m / s. The recovery release field is judged as follows: the PIR wake-up signal after the last test has recovered to the silence state and has lasted for more than the recovery stabilization time Treset. The Treset for vehicle heat source test is set to 15s, and the Treset for pet heat source test is set to 10s. The distance parameter lock field is judged as follows: the candidate equivalent distance d, candidate pixel set Ω, duty cycle u, lighting duration τ, and lighting order Q have been written into the current test record and have not been modified before generating the one-time test permission status record.

[0061] During each candidate equivalent distance test for vehicle heat source testing and pet heat source testing, the main controller 8 synchronously calculates the multi-residual anomaly identification record. The multi-residual anomaly identification record includes at least the lamp wall radiation residual R1, the cumulative heat dose residual R2, the ambient temperature residual R3, the background infrared residual R4, the dormant current residual R5, and the wake-up time synchronization residual R6. R1=|Ltarget-Lmeas| / Ltarget; R2=|Hmeas-Href| / Href; R3=|Ta-Ta,ref|; R4=|BIR-BIR,0| / BIR,0; R5=|Iidle-Iidle,ref| / Iidle,ref; R6=|twake,gpio-twake,current|. In one optional implementation, the over-limit thresholds of R1, R2, R3, R4, R5 and R6 can be set to 3%, 5%, 2°C, 5%, 20% and 100ms respectively; the specific thresholds can be adjusted according to the model of the low-power camera under test, the calibration results of the light wall component and the test accuracy requirements.

[0062] When R1 exceeds the limit but R3, R4, and R5 do not, the main controller 8 determines that the source of the anomaly is an abnormal radiation output of the lamp wall component 2, and recalibrates the far-infrared heating pixels corresponding to the R1 exceeding limit area; when R2 exceeds the limit but R1 does not, the main controller 8 determines that the source of the anomaly is an abnormal hot spot trajectory or cumulative thermal process, and recalculates the hot spot center trajectory, lighting sequence Q, and lighting duration τ; when R4 exceeds the limit, the main controller 8 determines that there is background infrared disturbance in the test dark box 1, stops the output of the lamp wall component 2, and re-executes baseline detection; when R5 exceeds the limit, the main controller 8 determines that the tested low-power camera 4 has not entered an effective low-power sleep state, and resends the sleep command; when R6 exceeds the limit, the main controller 8 resends the synchronization timestamp and verifies the sampling time reference. If the same anomaly source occurs 3 times consecutively, the main controller 8 marks the data corresponding to the candidate equivalent distance as abnormal invalid data and does not include it in the calculation of the trigger probability p(d).

[0063] After the equivalent distance tests for each candidate heat source (vehicle and pet) are completed, the main controller 8 performs logistic curve fitting on the trigger probability p(d) of the vehicle and pet heat sources respectively, and outputs the corresponding D50, D90, and D95 ​​trigger distance indices. The report generation module 10 generates a test report, which includes at least the model of the low-power camera under test, the test darkroom environment conditions, the specifications of the light wall components, the test parameters for the vehicle heat source and the pet heat source, the number of triggers and the trigger probability at each candidate equivalent distance, the mean response time, the standard deviation of the response time, the D50 / D90 / D95 values ​​for the vehicle heat source and the pet heat source, the pixel health status table, the number of abnormal far-infrared heating pixels, the neighbor replacement reconstruction record, the multi-disability abnormal identification record, the abnormal isolation result, the recovery action, and the reason for invalid data removal.

[0064] The above description is merely a preferred embodiment of the present invention and does not limit the scope of protection of the present invention. Any equivalent structural transformations or equivalent step substitutions made based on the concept of the present invention and the description and drawings should be included within the scope of protection of the present invention.

Claims

1. An automatic testing device for the PIR trigger distance of a low-power camera, characterized in that, It includes a test dark box, a light wall assembly, a positioning assembly for the low-power camera under test, an infrared feedback acquisition assembly, an environmental acquisition assembly, a device status acquisition assembly, a main controller, a memory, and a report generation module; The test chamber is a box-shaped structure with a closed test cavity. The inner wall of the test chamber is provided with a low-reflection, low-thermal-inertia lining layer to reduce the impact of external environmental thermal disturbance and thermal hysteresis of the inner wall of the test chamber on PIR triggering test. The first end of the test cavity is located on the side where the lamp wall assembly is located, and the second end of the test cavity is located on the side where the low-power camera positioning assembly under test is located. The lamp wall assembly is fixedly installed at the first end of the test cavity. The lamp wall assembly has an array emission surface facing the second end of the test cavity. The array emission surface is formed by M×N independently driveable far-infrared heating pixels arranged in a row and column matrix, where M and N are both integers greater than 1. The positioning component of the low-power camera under test is disposed at the second end of the test cavity. The positioning component of the low-power camera under test is used to fix the low-power camera under test and make the PIR window of the low-power camera under test face the array emission surface, and make the PIR detection center axis of the low-power camera under test pass through the geometric center of the array emission surface, or the lateral deviation and vertical deviation between the projection point of the PIR detection center axis on the array emission surface and the geometric center are not greater than half of the minimum center distance of adjacent far-infrared heating pixels and are not greater than 5mm. The infrared feedback acquisition component is installed on the top wall, side wall, or lateral unobstructed position of the lamp wall component of the test dark box. The detection field of view of the infrared feedback acquisition component covers the array emission surface and is used to acquire the actual radiation intensity and hot spot morphology of each far-infrared heating pixel. The environmental acquisition component is installed on the inner wall of the test chamber and located in the non-direct hot spot area between the lamp wall component and the positioning component of the low-power camera under test. It is used to collect environmental temperature, humidity, background infrared radiation and airflow data in the test chamber. The device status acquisition component is connected to or aligned with the status display area of ​​the low-power camera under test, and is used to acquire the PIR wake-up status and wake-up time of the low-power camera under test. The main controller is communicatively connected to the light wall assembly, the infrared feedback acquisition assembly, the environmental acquisition assembly, the equipment status acquisition assembly, the memory, and the report generation module. The main controller is configured to generate light wall control parameters based on the real heat source feature library and pixel radiation calibration data in the memory, and control the far-infrared heating pixel array to synthesize an equivalent moving heat source. The driving parameters of the far-infrared heating pixel are corrected in a closed loop based on the actual radiation intensity collected by the infrared feedback acquisition component. Based on the PIR wake-up status collected by the device status acquisition component, an adaptive search for trigger distance is performed, and the test results of PIR trigger distance and response time are output by the report generation module.

2. The automatic testing device for low-power camera PIR trigger distance according to claim 1, characterized in that, The light wall assembly includes a plate-shaped support frame, a pixel mounting plate, a drive circuit board, and multiple far-infrared heating pixels; The plate-shaped support frame is fixed on the inner wall of the first end of the test dark box. The pixel mounting plate is a flat structure and is fixed on the side of the plate-shaped support frame near the positioning component of the low-power camera under test. Multiple far-infrared heating pixels are mounted on the pixel mounting plate with equal or preset unequal row and column spacing. Each far-infrared heating pixel is a sheet or block of ceramic far-infrared heating element, controllable blackbody radiation unit or micro heating film. The front of each far-infrared heating pixel constitutes part of the array emission surface, and the back of each far-infrared heating pixel is electrically connected to the drive circuit board. The driving circuit board has an independent driving channel corresponding to each far-infrared heating pixel. The independent driving channel can adjust the PWM duty cycle, driving current, lighting duration and cooling interval of the corresponding far-infrared heating pixel respectively. The low-power camera positioning component under test includes a rectangular base, a vertical adjustment bracket, and an angle adjustment seat. The rectangular base is fixed to the bottom wall of the test dark box, the vertical adjustment bracket is vertically set on the rectangular base, and the angle adjustment seat is installed on the side of the vertical adjustment bracket near the light wall component. The angle adjustment base has a mounting surface for supporting or clamping the low-power camera under test. The mounting surface is oriented towards the array emission surface. The angle adjustment base can adjust the pitch angle and / or horizontal deflection angle of the low-power camera under test, so that the center of the PIR window of different models of low-power cameras under test is aligned with the array center of the light wall assembly.

3. The automatic testing device for the PIR trigger distance of a low-power camera according to claim 1, characterized in that, The infrared feedback acquisition component includes a thermal imager or a thermopile array; The infrared feedback acquisition component is mounted on the top or side wall of the test dark chamber by a fixed bracket. The detection axis of the infrared feedback acquisition component is inclined relative to the array emission surface of the lamp wall component and avoids the PIR detection center axis of the low-power camera under test, so as to acquire the radiation distribution of the array emission surface without obstructing the PIR detection field of view. The environmental acquisition component includes a temperature sensor, a humidity sensor, a background infrared radiation sensor, and an airflow sensor. The temperature sensor and humidity sensor are installed in the middle of the inner wall of the test chamber. The background infrared radiation sensor is positioned facing the non-heating area of ​​the inner wall of the test chamber. The airflow sensor is installed near the positioning component of the low-power camera under test and avoids the direct radiation area of ​​the light wall component. The device status acquisition component includes one or more of the following: serial port interface, USB interface, network log acquisition interface, GPIO acquisition interface, power current sampling unit and status indicator detection unit. The power current sampling unit is connected in series or coupled to the power supply line of the low-power camera under test, and the status indicator detection unit is located in front of or diagonally in front of the status indicator of the low-power camera under test. The memory stores a real heat source feature library, a pixel radiation calibration table, and a lamp wall control parameter database. The real heat source feature library stores the feature vector F of the real heat source when it passes through the PIR field of view according to the heat source category c, distance d, moving speed v, ambient temperature Ta, and moving direction θ. The feature vector F includes at least the target-background temperature difference ΔT, the equivalent hot spot area A, the hot spot center trajectory x(t), y(t), the thermal radiation peak value Rpeak, the rise time tr, the fall time tf, the PIR equivalent output peak value Vpeak, the effective trigger duration teff, and the noise baseline σ. The pixel radiation calibration table is established according to pixel number, driving parameters, ambient temperature, radiation intensity, heating time and cooling time; the lamp wall control parameter database establishes mapping records according to heat source type c, equivalent distance d, ambient temperature Ta, moving speed v, candidate pixel set Ω, duty cycle u, lighting duration τ and lighting order Q.

4. The automatic testing device for the PIR trigger distance of a low-power camera according to claim 3, characterized in that, The main controller is configured to perform distance-power-trajectory matching and dynamic heat source trajectory generation; After setting the heat source category c and the equivalent distance d, the main controller reads the corresponding reference feature Fref from the real heat source feature library and generates a candidate pixel set Ω based on the geometric coordinates of each far-infrared heating pixel on the array emission surface. The main controller solves for the time-varying driving vector U={uij(t)} of each far-infrared heating pixel to minimize the error between the simulated hot spot feature Fsim synthesized by the lamp wall assembly and the reference feature Fref; where uij(t) represents the driving duty cycle or normalized driving intensity of the far-infrared heating pixel in the i-th row and j-th column at time t, Fsim includes at least the simulated radiation peak value Rsim, the simulated hot spot area Asim, the simulated rise time tr,sim, and the simulated PIR equivalent output peak value Vsim, and Fref includes at least the reference radiation peak value Rref, the reference hot spot area Asim, the reference rise time tr,ref, and the reference PIR equivalent output peak value Vref; The minimum error is evaluated using an objective function E. E=wR·|Rsim-Rref| / Rref+wA·|Asim-Aref| / Aref+wt·|tr,sim-tr,ref| / tr,ref+wV·|Vsim-Vref| / Vref+λ·[1 / (|T|·|GΩ|)]·Σ_{t∈T}Σ_{((i,j),(p,q))∈GΩ}[(uij(t)-upq(t))²], where T is the set of discrete sampling times of the driving vector U, GΩ is the set of unordered pairs of adjacent pixels formed by the four neighborhood relationships of top, bottom, left, and right within the candidate pixel set Ω, |T| and |GΩ| are the number of elements in the corresponding set, wR, wA, wt and wV are non-negative weights, λ is the power smoothing coefficient of adjacent pixels, and upq(t) is the driving intensity of the far-infrared heating pixel in the p-th row and q-th column paired with the far-infrared heating pixel in the i-th row and j-th column at time t; The main controller also generates a driving vector U based on the power smoothing constraint of adjacent pixels to avoid sudden changes between adjacent far-infrared heating pixels that do not conform to the actual heat source movement law. When simulating a lateral moving heat source, the main controller determines the switching period Δt=s / v of adjacent hot spot windows based on the target moving speed v and the spacing s of adjacent far-infrared heating pixels, and sequentially changes the set of candidate pixels Ω to be lit along the lateral direction of the array output surface. When simulating approaching or moving away from a heat source, the main controller achieves distance variation by changing the radiation power level and hot spot area corresponding to the equivalent distance d; when simulating a vehicle or pet heat source, the main controller forms an equivalent heat source pattern that is different from a human body heat source by changing the hot spot height, hot spot width, temperature distribution, and trajectory curvature. The dynamic heat source trajectory satisfies the following relationship: uij(t)=u0(t)·exp{-[(xij-xc(t))² / (2σx²)+(yij-yc(t))² / (2σy²)]}; where uij(t) is the driving intensity of the far-infrared heating pixel in the i-th row and j-th column at time t, u0(t) is the center driving intensity at time t, xij and yij are the array coordinates of the far-infrared heating pixel in the i-th row and j-th column, xc(t) and yc(t) are the center coordinates of the hot spot at time t, and σx and σy are the lateral diffusion parameter and longitudinal diffusion parameter related to the heat source type and equivalent distance, respectively.

5. The automatic testing device for low-power camera PIR trigger distance according to claim 1, characterized in that, The main controller is configured to perform closed-loop radiation control, baseline detection, and outlier data removal. During the output of the equivalent heat source by the lamp wall assembly, the infrared feedback acquisition assembly acquires the actual radiation intensity Lmeas of the array output surface. The main controller compares the actual radiation intensity Lmeas with the target radiation intensity Ltarget to obtain the error e(t) = Ltarget(t) - Lmeas(t), and corrects the driving parameters of the corresponding far-infrared heating pixel based on the error e(t). When the absolute value of the error e(t) is less than the preset error threshold ε within m consecutive sampling periods, the main controller determines that the current equivalent heat source has reached a stable state, where m is an integer from 3 to 10, and ε is 1% to 3% of the target radiation intensity Ltarget; when the absolute value of the error e(t) exceeds the preset error threshold ε, the main controller records the corresponding far-infrared heating pixel as a calibration abnormal pixel and recalibrates or removes the test data corresponding to the far-infrared heating pixel. Before each round of testing, the main controller monitors whether the low-power camera under test generates a PIR wake-up event through the device status acquisition component when the light wall component does not output a heat source. If a PIR wake-up event is generated, it is determined that there is background disturbance or the low-power camera under test is abnormal, and the test is stopped from proceeding to the valid test.

6. An automatic testing method for the PIR trigger distance of a low-power camera, characterized in that, The method is performed using the automatic testing device for the low-power camera PIR trigger distance according to any one of claims 1 to 5, and includes the following steps: S1. Install the low-power camera under test on the positioning component of the low-power camera under test, so that the PIR window of the low-power camera under test faces the array emission surface of the light wall component, and align the PIR detection center axis with the geometric center of the array emission surface. S2. Read or build a real heat source feature library, which stores the feature vector F of a real heat source when it passes through the PIR field of view according to the heat source category c, distance d, moving speed v, ambient temperature Ta and moving direction θ. S3. Perform radiation calibration on each far-infrared heating pixel in the light wall assembly, establish a pixel radiation calibration table, and generate a light wall control parameter database based on the real heat source feature library and the pixel radiation calibration table. S4. Based on the target heat source type, target equivalent distance, target moving speed and current ambient temperature, read the corresponding candidate pixel set Ω, duty cycle u, lighting duration τ and lighting sequence Q from the light wall control parameter database, and control the light wall component to output the equivalent moving heat source. S5. Control or wait for the low-power camera under test to enter low-power sleep mode, and confirm through the device status acquisition component that the low-power camera under test only retains the wake-up function required for PIR detection. S6. At each candidate equivalent distance, control the lamp wall component to repeatedly output the equivalent moving heat source, and use the infrared feedback acquisition component to perform closed-loop correction on the actual radiation intensity of the lamp wall component. S7. Collect the PIR wake-up status and wake-up time of the low-power camera under test through the device status acquisition component, and calculate the trigger probability and response time under the candidate equivalent distance. S8. Perform adaptive search for trigger distance based on the trigger probability and response time under different candidate equivalent distances to determine the PIR trigger distance of the low-power camera under test, and generate a test report containing trigger distance, response time, repeatability statistics and abnormal records.

7. The automatic testing method for the PIR trigger distance of a low-power camera according to claim 6, characterized in that, The adaptive search for trigger distance includes a coarse scan stage, a boundary search stage, and a probability fitting stage. During the coarse scan phase, equivalent moving heat sources are output sequentially according to a preset distance sequence from far to near: dmax, dmax-Δd, dmax-2Δd, ..., dmin. The trigger boundary interval [dnear, dfar] is determined based on whether the heat source is triggered. Here, dnear is the closer distance that meets the stable triggering condition, and dfar is the farther distance that is not triggered or has insufficient triggering probability. Furthermore, dnear is less than dfar. During the boundary search phase, a binary search or a staircase search is performed within the trigger boundary interval [dnear, dfar]. For each candidate equivalent distance d, the test is repeated k times, the number of triggers m is recorded, the trigger probability p(d) = m / k is calculated, and the mean response time μt and the standard deviation of response time σt are recorded. When p(d) ≥ Pth and μt ≤ τth, the candidate equivalent distance d is determined to meet the triggering requirements, where Pth is a preset trigger probability threshold and τth is a preset response time threshold. During the probability fitting stage, the trigger probability p(d) is fitted with a logistic curve, and the fitting relationship is p(d)=1 / [1+exp(a·(db))], where a is the trigger boundary slope parameter and b is the distance parameter near the trigger probability of 50%. When a is greater than 0, the distance index D(P0) with the trigger probability target value P0 is calculated according to D(P0)=b+ln(1 / P0-1) / a, where P0 is a preset probability target value between 0 and 1. The main controller outputs D50, D90 and / or D95 as PIR trigger distance indexes based on the fitted curve and generates confidence intervals. D50, D90 and D95 ​​correspond to D(0.50), D(0.90) and D(0.95) respectively. The response time is calculated as tresponse = twake - ton, where ton is the moment when the light wall component starts outputting the equivalent moving heat source, and twake is the moment when the low-power camera under test generates a PIR wake-up event. The ton and twake are recorded by the same master clock, or by the master controller sending a synchronization timestamp to the light wall component, the device status acquisition component, and the interface of the low-power camera under test.

8. The automatic testing method for the PIR trigger distance of a low-power camera according to claim 7, characterized in that, Before each candidate equivalent distance d is tested, the main controller first establishes a one-time test permission status record corresponding to the candidate equivalent distance d. The one-time test permission status record is used to characterize that the light wall component, the low-power camera under test, the test dark box environment, and the distance control parameters are all in the allowed test state for this candidate equivalent distance test. The one-time test permit status record is generated based on at least the light wall stability field, PIR silence field, environmental stability field, recovery release field, and distance parameter lock field. Wherein, the lamp wall stability field is set to allow when the absolute value of the error between the actual radiation intensity Lmeas and the target radiation intensity Ltarget is less than a preset error threshold ε within m consecutive sampling periods, where m is an integer from 3 to 10, and ε is 1% to 3% of the target radiation intensity; The PIR silence field is enabled when the GPIO acquisition interface does not detect a PIR wake-up transition, the power current sampling unit detects that the low-power camera under test is in the sleep current range, and the status indicator detection unit does not detect a change in the wake-up indication. The environmental stability field is enabled when the rate of change of ambient temperature in the test chamber is less than 0.2℃ / min, the rate of change of background infrared radiation is less than a preset background threshold, and the airflow speed is less than 0.2m / s; the recovery release field is enabled when the PIR wake-up signal after the last test returns to the silent state and continues for more than the recovery stabilization time Treset. The distance parameter locking field is set to allowed if it has not been modified after the candidate equivalent distance d, candidate cell set Ω, duty cycle u, lighting duration τ, and lighting order Q are written into the current test record; When the light wall stability field, PIR silence field, environmental stability field, recovery release field, and distance parameter lock field are all allowed, the main controller generates the one-time test permission status record and controls the light wall component to output an equivalent moving heat source; when any field is not allowed, the main controller does not output an equivalent moving heat source and records the reason for non-permission; after each equivalent moving heat source output ends, the one-time test permission status record becomes invalid, and a new one-time test permission status record is generated before the next test.

9. The automatic testing method for the PIR trigger distance of a low-power camera according to claim 7, characterized in that, Before the output of the equivalent moving heat source of the light wall component, the main controller establishes a pixel health status table of far-infrared heating pixels and performs neighborhood replacement reconstruction on abnormal far-infrared heating pixels based on the pixel health status table. The pixel health status table stores the pixel number, row coordinates, column coordinates, radiation amplitude deviation, heating time deviation, cooling time deviation, and health status of each far-infrared heating pixel according to the row and column coordinates of the far-infrared heating pixels on the array emitting surface. When establishing the pixel health status table, the main controller applies low duty cycle, medium duty cycle, and high duty cycle drive signals to each far-infrared heating pixel, and the infrared feedback acquisition component acquires the actual radiation intensity Lij, heating time tup,ij, and cooling time tdown,ij of the far-infrared heating pixel under different duty cycles. The radiation amplitude deviation δLij is calculated according to δLij=|Lij-Lref,ij| / Lref,ij, where Lref,ij is the reference radiation intensity of the far-infrared heating pixel in the i-th row and j-th column of the pixel radiation calibration table. The heating time deviation δtup,ij is calculated according to δtup,ij=|tup,ij-tup,ref,ij| / tup,ref,ij, where tup,ref,ij is the reference heating time of the far-infrared heating pixel in the i-th row and j-th column of the pixel radiation calibration table. The cooling time deviation δtdown,ij is calculated according to δtdown,ij=|tdown,ij-tdown,ref,ij| / tdown,ref,ij, where tdown,ref,ij is the reference cooling time of the far-infrared heating pixel in the i-th row and j-th column of the pixel radiation calibration table. When δLij is greater than 10%, δtup,ij is greater than 15%, or δtdown,ij is greater than 15%, the main controller will mark the corresponding far-infrared heating pixel as an abnormal pixel. The health status includes one of the following: normal, insufficient radiation, heating lag, cooling lag, and failure. When the candidate pixel set Ω contains an abnormal pixel, the main controller selects at least two far-infrared heating pixels with normal health status from above, below, left, right and diagonally adjacent positions of the abnormal pixel as a replacement pixel set Ωr, and assigns compensation weights according to the distance between the replacement pixels and the abnormal pixels, so that the replacement pixel set Ωr jointly outputs the target radiation intensity originally corresponding to the abnormal pixel. After the neighborhood replacement reconstruction is completed, the main controller re-acquires the actual hot spot output by the lamp wall component through the infrared feedback acquisition component, and determines whether the compensated hot spot center deviation, hot spot area deviation, and cumulative thermal dose spectrum deviation are all less than the corresponding thresholds. The cumulative thermal dose spectrum deviation ΔHmap is calculated according to ΔHmap=Σ|Hmeas(i,j)-Href(i,j)| / ΣHref(i,j), where Hmeas(i,j) is the measured cumulative thermal dose of the far-infrared heating pixel in the i-th row and j-th column during one equivalent moving heat source output process, and Href(i,j) is the corresponding reference cumulative thermal dose. If all three deviations are less than the corresponding thresholds, the candidate equivalent distance test is allowed to continue. If any deviation exceeds the corresponding threshold, the candidate equivalent distance test is marked as lamp wall reconstruction failure data and is not included in the calculation of the trigger probability p(d).

10. The automatic testing method for the PIR trigger distance of a low-power camera according to claim 7, characterized in that, During each test of the candidate equivalent distance d, the main controller synchronously calculates the multi-damage anomaly identification record and determines the source of the anomaly and the recovery action based on the multi-damage anomaly identification record; The multi-residual abnormality identification record includes at least the lamp wall radiation residual R1, cumulative thermal dose residual R2, ambient temperature residual R3, background infrared residual R4, dormant current residual R5, and wake-up time synchronization residual R6. Where R1 = |Ltarget - Lmeas| / Ltarget, Ltarget is the target radiation intensity, and Lmeas is the actual radiation intensity; R2 = |Hmeas - Href| / Href, Hmeas is the measured cumulative heat dose collected by the infrared feedback acquisition component during one equivalent moving heat source output process, and Href is the reference cumulative heat dose in the real heat source feature library under the corresponding heat source type, equivalent distance, and moving speed; R3 = |Ta - Ta,ref|, Ta ​​is the current ambient temperature in the test dark box, and Ta,ref is the reference ambient temperature in the real heat source feature library corresponding to the test conditions; R4 = |BIR - BIR,0| / BIR,0, where BIR is the current background infrared radiation and BIR,0 is the baseline background infrared radiation when the light wall component does not output an equivalent moving heat source; R5 = |Iidle - Iidle,ref| / Iidle,ref, where Iidle is the current sleep current of the low-power camera under test and Iidle,ref is the reference sleep current of the low-power camera under test; R6 = |twake,gpio - twake,current|, where twake,gpio is the PIR wake-up time recorded by the GPIO acquisition interface and twake,current is the current surge wake-up time recorded by the power supply current sampling unit. When R1 exceeds the limit but R3, R4, and R5 do not, the main controller determines that the source of the abnormality is an abnormal radiation output of the lamp wall component and re-executes the radiation calibration of the corresponding far-infrared heating pixel; when R2 exceeds the limit but R1 does not, the main controller determines that the source of the abnormality is an abnormality of the hot spot trajectory or the cumulative thermal process and recalculates the hot spot center trajectory, lighting sequence Q, and lighting duration τ. When R4 exceeds the limit, the main controller determines that there is background infrared disturbance in the test dark box and suspends the test until the background infrared radiation recovers to the baseline allowable range before re-performing baseline detection; when R5 exceeds the limit, the main controller determines that the low-power camera under test has not entered a valid low-power sleep state and resends the sleep command or outputs a device sleep abnormality prompt; when R6 exceeds the limit, the main controller determines that the device status acquisition component has a wake-up time acquisition synchronization error and resends the synchronization timestamp to the light wall component, the device status acquisition component, and the interface of the low-power camera under test. After the above recovery action, the main controller re-executes the test of the corresponding candidate equivalent distance d; if the same source of anomaly occurs consecutively a preset number of times, the data corresponding to the candidate equivalent distance d is marked as abnormal invalid data, and the source of anomaly, residual value and recovery action are recorded in the test report.

Images