Kalman filtering method and related apparatus

Abstract

The application provides a Kalman filtering method and related device, the method comprising: constructing a three-dimensional Kalman filtering coefficient table, determining a stability level of current target detection according to appearance feature cumulative deviation, distance cumulative deviation and speed cumulative deviation; determining a target Kalman filtering intensity value of the current target at the current time according to the stability level and the three-dimensional Kalman filtering coefficient table; and updating the state of the current target according to the Kalman filtering intensity value. The effective information of the target is converted into a gain coefficient of the Kalman coefficient table, the Kalman filter is converted into a form of a fixed coefficient table, the real-time three-dimensional index is determined, the table lookup method is used to replace the matrix inversion operation each time, and the calculation efficiency is improved.

Description

Technical Field

[0001] This application belongs to the field of signal processing, specifically relating to a Kalman filtering method and related apparatus. Background Technology

[0002] Currently, there are two main problems with the use of Kalman filters. First, Kalman filters involve multiple matrix inversion operations. If a single target has many parameters, the filter matrix has a high dimension, and the number of targets to be filtered is large, the program's execution time will increase significantly. Second, for basic Kalman algorithms or extended Kalman algorithms that do not change the noise, when a state is maintained for a long time, the output of the Kalman filter algorithm remains stable, but its steady state may not achieve the desired filtering strength. External interference is needed to compare the changes in noise, but using an adaptive Kalman filter algorithm introduces even more time consumption. Summary of the Invention

[0003] This application provides a Kalman filtering method and related apparatus. By converting the effective information of the target into gain coefficients of the Kalman coefficient table, the Kalman filter is converted into a coefficient table with fixed coefficients. By determining the three-dimensional index in real time, the lookup table method is used to replace each matrix inversion operation, which helps to improve computational efficiency.

[0004] In a first aspect, embodiments of this application provide a Kalman filtering method, the method comprising:

[0005] Obtain basic filter parameter information;

[0006] Construct a three-dimensional Kalman filter coefficient table based on the basic filter parameter information;

[0007] Obtain observation information of the current frame detection box for the current target, as well as historical distance data and historical velocity data of the current target;

[0008] The cumulative deviation of appearance features is determined based on the observation information of the detection box in the current frame;

[0009] Based on the historical distance data and historical speed data, determine the cumulative distance deviation and cumulative speed deviation;

[0010] The stability level of the current target detection is determined based on the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the velocity.

[0011] Based on the stability level and the three-dimensional Kalman filter coefficient table, determine the target Kalman filter strength value of the current target at the current moment;

[0012] The current target's state is updated based on the Kalman filter strength value.

[0013] In one possible example, the basic filtering parameter information includes the current target's tracking distance range, preset Kalman noise, and operating condition information; the step of constructing a filtering parameter table based on the basic filtering parameter information includes:

[0014] Based on the tracking distance range, multiple noise levels are set, and each noise level corresponds to a set of Kalman filter noise parameters;

[0015] A two-dimensional Kalman parameter table is determined based on the preset Kalman noise and the Kalman filter noise parameters corresponding to each noise level.

[0016] Based on the operating condition information, multiple Kalman filter strength levels are determined, and the filtering factor for each Kalman filter strength level is generated.

[0017] Based on the filter ratio and the two-dimensional Kalman parameter table, determine the three-dimensional Kalman filter coefficient table.

[0018] In one possible example, the observation information includes the width ratio of the current frame width to the previous frame width and the height ratio of the current frame height to the previous frame height of the current frame detection box; the cumulative deviation of the appearance features includes the cumulative deviation of the width ratio and the cumulative deviation of the height ratio; determining the cumulative deviation of the appearance features based on the observation information of the current frame detection box includes:

[0019] Acquire historical observation information, which includes historical width ratio data and historical height ratio data of the detection box. The historical width ratio data includes multiple historical width ratios, and the historical height ratio data includes multiple historical height ratios.

[0020] By fitting the historical width ratio data, a first straight line is obtained;

[0021] By fitting the historical height ratio data, a second straight line is obtained;

[0022] Determine the cumulative deviation of the current frame width ratio from the first straight line;

[0023] Determine the cumulative deviation of the current frame height ratio from the height ratio of the second straight line.

[0024] In one possible example, determining the cumulative distance deviation and cumulative speed deviation based on the historical distance data and historical speed data includes:

[0025] By fitting the historical distance data, a first curve is obtained;

[0026] By fitting the historical velocity data, a second curve is obtained;

[0027] Determine the cumulative distance deviation of the current frame from the first curve;

[0028] Determine the cumulative velocity deviation of the current frame rate from the second curve.

[0029] In one possible example, the stability level of the current target detection is determined based on the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the velocity, including:

[0030] Based on the observation information, determine the occlusion attributes;

[0031] Based on the occlusion attributes, determine the fusion weights of the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the velocity;

[0032] The feature fluctuation fusion value is determined based on the fusion weight, the occlusion attribute, the cumulative deviation of appearance features, the cumulative deviation of distance, and the cumulative deviation of velocity.

[0033] Based on the fusion value of the characteristic fluctuations, the stability level of the current target detection is determined.

[0034] In one possible example, the occlusion attribute is negatively correlated with the fusion weight of the cumulative deviation of the appearance feature, and positively correlated with the fusion weight of the cumulative distance deviation and the cumulative velocity deviation.

[0035] In one possible example, the target Kalman filter strength value of the current target at the current moment is determined based on the stability level and the three-dimensional Kalman filter coefficient table, including:

[0036] Based on the stability level, determine the target Kalman filter strength level;

[0037] Based on the observation information, determine the time index and target noise level.

[0038] The target Kalman filter strength value of the current target at the current time is obtained by querying the three-dimensional Kalman filter coefficient table based on the target Kalman filter strength level, the time index, and the target noise level.

[0039] Secondly, embodiments of this application provide a Kalman filter device, which includes an acquisition unit, a construction unit, a determination unit, and an update unit; wherein,

[0040] The acquisition unit is used to acquire basic filter parameter information;

[0041] The construction unit is used to construct a three-dimensional Kalman filter coefficient table based on the basic filter parameter information.

[0042] The acquisition unit is also used to acquire observation information of the current frame detection box for the current target, and to acquire historical distance data and historical velocity data of the current target;

[0043] The determining unit is used to determine the cumulative deviation of appearance features based on the observation information of the detection box in the current frame;

[0044] The determining unit is further configured to determine the cumulative distance deviation and the cumulative speed deviation based on the historical distance data and the historical speed data;

[0045] The determining unit is further configured to determine the stability level of the current target detection based on the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the velocity.

[0046] The determining unit is further configured to determine the target Kalman filter strength value of the current target at the current moment based on the stability level and the three-dimensional Kalman filter coefficient table;

[0047] The update unit is used to update the state of the current target based on the Kalman filter intensity value.

[0048] A third aspect of this application provides an electronic device including: a processor and a memory; and one or more programs stored in the memory and configured to be executed by the processor, the programs including instructions for some or all of the steps as described in the first aspect.

[0049] A fourth aspect of this application provides a computer-readable storage medium for storing a computer program that causes a computer to perform some or all of the steps described in the first aspect of this application.

[0050] A fifth aspect of this application provides a computer program product, comprising a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps described in the first aspect of this application. This computer program product may be a software installation package.

[0051] As can be seen, in this embodiment, the basic filtering parameter information is first obtained. Then, a three-dimensional Kalman filter coefficient table is constructed based on the basic filtering parameter information. Next, the observation information of the current frame detection box for the current target, as well as the historical distance data and historical velocity data of the current target, are obtained. Further, the cumulative deviation of appearance features is determined based on the observation information of the current frame detection box. Then, based on the historical distance data and historical velocity data, the cumulative distance deviation and cumulative velocity deviation are determined. Next, based on the cumulative deviation of appearance features, cumulative distance deviation, and cumulative velocity deviation, the stability level of the current target detection is determined. Then, based on the stability level and the three-dimensional Kalman filter coefficient table, the target Kalman filter intensity value of the current target at the current moment is determined. Finally, the state of the current target is updated based on the Kalman filter intensity value. By converting the effective information of the target into gain coefficients for the Kalman coefficient table, and converting the Kalman filter into a coefficient table with fixed coefficients, and by determining the three-dimensional index in real time, a lookup table method is used to replace each matrix inversion operation, which helps to improve computational efficiency. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 This is a schematic flowchart of a Kalman filtering method provided in an embodiment of this application;

[0054] Figure 2 This is a schematic diagram of a process for constructing a filter parameter table provided in an embodiment of this application;

[0055] Figure 3 This is a schematic diagram of a process for determining cumulative deviation provided in an embodiment of this application;

[0056] Figure 4 This is a schematic diagram of another process for determining cumulative deviation provided in an embodiment of this application;

[0057] Figure 5 This is a schematic diagram of a process for determining a stability level provided in an embodiment of this application;

[0058] Figure 6 This is a schematic diagram of a process for determining the Kalman filter strength value provided in an embodiment of this application;

[0059] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application;

[0060] Figure 8 This is a block diagram of the functional units of a Kalman filter device provided in an embodiment of this application. Detailed Implementation

[0061] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0062] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0063] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0064] In the embodiments of this application, "and / or" describes the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent the following three situations: A exists alone; A and B exist simultaneously; B exists alone. Among them, A and B can be singular or plural.

[0065] In this embodiment, the symbol " / " can indicate that the preceding and following objects are in an "or" relationship. Alternatively, the symbol " / " can also represent a division sign, i.e., performing a division operation. For example, A / B can mean A divided by B.

[0066] In the embodiments of this application, "at least one item" or its similar expression refers to any combination of these items, including any combination of a single item or a plurality of items. "One or more" means one or more, while "multiple" means two or more. For example, "at least one item" of a, b, or c can represent the following seven cases: a, b, c; a and b; a and c; b and c; a, b, and c. Each of a, b, and c can be an element or a set containing one or more elements.

[0067] In the embodiments of this application, "equal to" can be used with "greater than" and is applicable to technical solutions used when "greater than" is used; it can also be used with "less than" and is applicable to technical solutions used when "less than" is used. When "equal to" is used with "greater than", it is not used with "less than"; when "equal to" is used with "less than", it is not used with "greater than".

[0068] The electronic device in this application embodiment is a device with wireless communication capabilities, and may be referred to as a terminal, user equipment (UE), mobile station (MS), mobile terminal (MT), access terminal device, vehicle-mounted terminal device, industrial control terminal device, UE unit, UE station, mobile station, remote station, remote terminal device, mobile device, UE terminal device, wireless communication device, UE agent, or UE device, etc. The terminal device can be fixed or mobile. It should be noted that the terminal device can support at least one wireless communication technology, such as LTE, New Radio (NR), Wideband Code Division Multiple Access (WCDMA), etc. For example, terminal devices can be mobile phones, tablets, desktop computers, laptops, all-in-one computers, in-vehicle terminals, virtual reality (VR) terminal devices, augmented reality (AR) terminal devices, wireless terminals in industrial control, wireless terminals in self-driving vehicles, wireless terminals in remote medical surgery, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, cellular phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), handheld devices with wireless communication capabilities, electronic devices or other processing devices connected to a wireless modem, wearable devices, terminal devices in future mobile communication networks, or terminal devices in future evolved public land mobile networks (PLMNs), etc. Electronic devices can be electronic rearview mirror devices.

[0069] Please see Figure 1 , Figure 1 This is a flowchart illustrating a Kalman filtering method provided in an embodiment of this application. The method includes:

[0070] Step S101: Obtain basic filter parameter information.

[0071] The basic filtering parameter information includes the preset target tracking distance range, the preset Kalman noise, and the target's actual operating conditions.

[0072] Step S102: Construct a three-dimensional Kalman filter coefficient table based on the basic filter parameter information.

[0073] Step S103: Obtain the observation information of the current frame detection box for the current target, as well as the historical distance data and historical velocity data of the current target.

[0074] The observation information includes at least one of the following: distance to the current target, speed, occlusion attributes, and width and height of the detection box.

[0075] The historical distance data includes multiple distances of the current target in multiple frames within a preset time period, and the historical speed data includes multiple speeds of the current target in multiple frames within a preset time period.

[0076] Step S104: Determine the cumulative deviation of appearance features based on the observation information of the detection box in the current frame.

[0077] The cumulative deviation of appearance features includes the cumulative deviation of width ratio and the cumulative deviation of height ratio.

[0078] Step S105: Determine the cumulative distance deviation and cumulative speed deviation based on the historical distance data and historical speed data.

[0079] Among them, the cumulative distance deviation is determined based on historical distance data, and the cumulative speed deviation is determined based on historical speed data.

[0080] Step S106: Determine the stability level of the current target detection based on the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the velocity.

[0081] The real-time stability of the current target detection is determined based on the cumulative deviation of appearance features, cumulative deviation of distance, and cumulative deviation of velocity, and then the filtering intensity is dynamically selected.

[0082] Step S107: Determine the target Kalman filter strength value of the current target at the current moment based on the stability level and the three-dimensional Kalman filter coefficient table.

[0083] Specifically, the target Kalman filter strength value is obtained by querying the three-dimensional Kalman filter coefficient table based on the stability level.

[0084] Step S108: Update the current target's state based on the Kalman filter strength value.

[0085] The filtering can be completed by performing a standard state update equation based on the Kalman filter strength value.

[0086] As can be seen, in this embodiment, the basic filtering parameter information is first obtained. Then, a three-dimensional Kalman filter coefficient table is constructed based on the basic filtering parameter information. Next, the observation information of the current frame detection box for the current target, as well as the historical distance data and historical velocity data of the current target, are obtained. Further, the cumulative deviation of appearance features is determined based on the observation information of the current frame detection box. Then, based on the historical distance data and historical velocity data, the cumulative distance deviation and cumulative velocity deviation are determined. Next, based on the cumulative deviation of appearance features, cumulative distance deviation, and cumulative velocity deviation, the stability level of the current target detection is determined. Then, based on the stability level and the three-dimensional Kalman filter coefficient table, the target Kalman filter intensity value of the current target at the current moment is determined. Finally, the state of the current target is updated based on the Kalman filter intensity value. By converting the effective information of the target into gain coefficients for the Kalman coefficient table, and converting the Kalman filter into a coefficient table with fixed coefficients, and by determining the three-dimensional index in real time, a lookup table method is used to replace each matrix inversion operation, which helps to improve computational efficiency.

[0087] Please see Figure 2 , Figure 2 This is a flowchart illustrating the construction of a filter parameter table according to an embodiment of this application. The basic filter parameter information includes the tracking distance range of the current target, preset Kalman noise, and operating condition information. The construction of the filter parameter table based on the basic filter parameter information includes:

[0088] Step S201: Based on the tracking distance range, set multiple noise levels, with each noise level corresponding to a set of Kalman filter noise parameters.

[0089] Based on the target tracking distance range, n different levels of process noise and observation noise are preset. There is a preset correspondence between the target tracking distance range and the noise level, and each level corresponds to a specific set of Kalman filter noise parameters.

[0090] Step S202: Determine the two-dimensional Kalman parameter table based on the preset Kalman noise and the Kalman filter noise parameters corresponding to each noise level.

[0091] Specifically, based on the preset Kalman noise, for each level of Kalman filter noise parameters, the complete gain sequence of the Kalman filter from the initial time to convergence is pre-calculated, obtaining key parameters such as Kalman gain and state covariance at m time points. This forms a... Two-dimensional Kalman parameter table.

[0092] Step S203: Based on the operating condition information, determine multiple Kalman filter strength levels and generate the filtering factor for each Kalman filter strength level.

[0093] Based on actual operating conditions, the Kalman filter intensity levels are divided, and a fixed filtering factor is generated for each level. Key parameters such as Kalman gain and state covariance in the generated two-dimensional Kalman parameter table are successively multiplied by the intensity factor to ultimately generate a... This table provides a 3D Kalman filter coefficient table, which includes Kalman filter intensity values. These values ​​are derived by multiplying key parameters such as Kalman gain and state covariance by an intensity factor. The table can be generated offline and includes filtering strategies for all preset operating scenarios.

[0094] As can be seen, in this example, by converting the target's effective information into gain coefficients of the Kalman coefficient table, the Kalman filter is converted into a coefficient table with fixed coefficients, which facilitates subsequent real-time querying and improves computational efficiency.

[0095] Please see Figure 3 , Figure 3 This is a flowchart illustrating a method for determining cumulative deviation, provided in an embodiment of this application. The observation information includes the width ratio of the current frame detection box to the width of the previous frame and the height ratio of the current frame height to the height of the previous frame. The cumulative deviation of the appearance feature includes the cumulative deviation of the width ratio and the cumulative deviation of the height ratio. The method for determining the cumulative deviation of the appearance feature based on the observation information of the current frame detection box may include the following steps:

[0096] Step S301: Obtain historical observation information, which includes historical width ratio data and historical height ratio data of the detection box. The historical width ratio data includes multiple historical width ratios, and the historical height ratio data includes multiple historical height ratios.

[0097] The historical width ratio data includes multiple historical width ratios across multiple frames within a preset time period, and the historical height ratio data includes multiple historical height ratios across multiple frames within a preset time period.

[0098] Step S302: Fit the historical width ratio data to obtain the first straight line.

[0099] The equation of the first straight line is obtained by fitting the historical width ratio data using the least squares method.

[0100] Step S303: Fit the historical height ratio data to obtain a second straight line.

[0101] The equation of the second straight line was obtained by fitting historical height data using the least squares method.

[0102] Step S304: Determine the cumulative deviation of the width ratio of the current frame from the first straight line.

[0103] Step S305: Determine the cumulative deviation of the height ratio of the current frame from the height ratio of the second straight line.

[0104] As can be seen in this example, historical data can be fitted to a straight line, and the cumulative deviation of the width ratio and the cumulative deviation of the height ratio can be calculated, which helps to improve the accuracy of the cumulative deviation calculation.

[0105] Please see Figure 4 In determining the cumulative distance deviation and cumulative speed deviation based on the historical distance data and historical speed data, the method includes the following steps:

[0106] Step S401: Fit the historical distance data to obtain a first curve;

[0107] Among them, the least squares method is used to fit the historical distance data to obtain the equation of the quadratic curve, that is, the first curve.

[0108] Step S402: Fit the historical speed data to obtain a second curve;

[0109] Among them, the least squares method is used to fit the historical speed data to obtain the equation of the quadratic curve, that is, the second curve.

[0110] Step S403: Determine the cumulative distance deviation of the current frame from the first curve;

[0111] Step S404: Determine the cumulative velocity deviation of the current frame velocity from the second curve.

[0112] As can be seen in this example, historical data can be fitted to a curve, and the cumulative distance deviation and cumulative speed deviation can be calculated, which helps to improve the accuracy of the cumulative deviation calculation.

[0113] Please see Figure 5 In determining the stability level of the current target detection based on the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the velocity, the above method may include the following steps:

[0114] Step S501: Determine the occlusion attribute based on the observation information;

[0115] Occlusion attributes quantify the degree, type, and area of ​​occlusion of a target by other objects and environmental elements in the scene. Types include hard occlusion (occlusion by solid objects, such as a vehicle being obscured by another car or tree, or a pedestrian being obscured by a lamppost, making the features completely invisible) and soft occlusion (occlusion by semi-transparent or blurred elements, such as rain-soaked glass, fog, or shadows, where some features can be extracted). Regions can be divided into core and non-core regions, which can be manually defined. Core regions include, for example, vehicle license plates, headlights, and corner points; pedestrian heads and torsos; and non-motorized vehicle handlebars and wheels.

[0116] For example, based on the degree, type, and area of ​​occlusion of the target by other objects and environmental elements in the scene, occlusion attributes are divided into four levels: no occlusion, slight occlusion, partial occlusion, and severe occlusion. No occlusion means that the core feature area of ​​the target is 100% visible, with no soft occlusion and no hard occlusion. Slight occlusion means that only non-core areas are softly occluded or a small area of ​​non-core areas is hardly occluded. Partial occlusion means that a local core area is hardly occluded or the occlusion rate of non-core areas exceeds 40%. Severe occlusion means that the core area is hardly occluded over a large area, with the overall occlusion rate exceeding 60%.

[0117] Step S502: Determine the fusion weights of the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the speed based on the occlusion attributes.

[0118] Specifically, the occlusion attribute is negatively correlated with the fusion weight of cumulative appearance feature deviation, while the occlusion attribute is positively correlated with the fusion weight of cumulative distance deviation and cumulative velocity deviation. That is, the more severe the occlusion, the smaller the weight set corresponding to the cumulative appearance feature deviation.

[0119] The sum of the fusion weights for cumulative distance deviation, cumulative velocity deviation, cumulative width ratio deviation, and cumulative height ratio deviation is 1.

[0120] Specifically, based on the occlusion attribute, a preset set of occlusion coefficients is queried to obtain the target occlusion coefficient. The set of occlusion coefficients includes the correspondence between occlusion attributes and occlusion coefficients, and the occlusion attributes and occlusion coefficients are positively correlated.

[0121] The formula for the characteristic fluctuation fusion value is as follows:

[0122] S=O+(a×E1+ b×E2+ c×E3+ d×E4).

[0123] Where S represents the feature fluctuation fusion value, E1 represents the cumulative deviation of the width ratio, a represents the fusion weight corresponding to the cumulative deviation of the width ratio, E2 represents the cumulative deviation of the height ratio, b represents the fusion weight corresponding to the cumulative deviation of the height ratio, E3 represents the cumulative deviation of the distance, c represents the fusion weight corresponding to the cumulative deviation of the distance, E4 represents the cumulative deviation of the velocity, d represents the fusion weight corresponding to the cumulative deviation of the velocity, and O represents the occlusion factor. The more severe the occlusion, the larger the occlusion factor; the less occlusion, the smaller the occlusion factor.

[0124] Step S503: Determine the feature fluctuation fusion value based on the fusion weight, the occlusion attribute, the cumulative deviation of appearance features, the cumulative deviation of distance, and the cumulative deviation of velocity;

[0125] Step S504: Determine the stability level of the current target detection based on the feature fluctuation fusion value.

[0126] Among them, the characteristic fluctuation fusion value is negatively correlated with the stability level, that is, the larger the characteristic fluctuation fusion value, the lower the stability level, and the smaller the characteristic fluctuation fusion value, the higher the stability level.

[0127] This system allows for the storage of a set of stability levels, which includes the correspondence between feature fluctuation fusion values ​​and stability levels. The stability level of the current target detection can be obtained by querying the stability level set using the feature fluctuation fusion values.

[0128] As can be seen, in this example, the stability of target detection at the current moment can be determined by combining the target's occlusion attributes with four cumulative biases, which helps to improve the accuracy of stability level determination.

[0129] Please see Figure 6 In determining the target Kalman filter strength value of the current target at the current time based on the stability level and the three-dimensional Kalman filter coefficient table, the above method may include the following steps:

[0130] Step S601: Determine the target Kalman filter strength level based on the stability level;

[0131] The correspondence between stability level and Kalman filter strength level can be preset. A set of Kalman filter strength levels can be stored in advance. The set of Kalman filter strength levels includes the correspondence between stability level and Kalman filter strength level. The target Kalman filter strength level can be obtained by querying the set of Kalman filter strength levels for stability level.

[0132] Step S602: Determine the time index and target noise level based on the observation information;

[0133] Among them, the time index of the current target can be determined in real time, and the corresponding target noise level can be determined in real time based on the tracking distance range of the current target. The time index is used to characterize the tracking time of the observed target.

[0134] Step S603: Query the three-dimensional Kalman filter coefficient table according to the target Kalman filter strength level, the time index, and the target noise level to obtain the target Kalman filter strength value of the current target at the current time.

[0135] The target Kalman filter strength level, time index, and target noise level are three indices for querying the three-dimensional Kalman filter coefficient table. By querying the three-dimensional Kalman filter coefficient table using these three indices, a unique corresponding Kalman filter strength value can be obtained.

[0136] As can be seen, in this example, by converting the target's effective information into gain coefficients of the Kalman coefficient table, and converting the Kalman filter into a coefficient table with fixed coefficients, and by determining the three-dimensional index in real time, the lookup table method is used to replace each matrix inversion operation, which helps to improve computational efficiency.

[0137] Please see Figure 7 , Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application, such as... Figure 7 As shown, the electronic device includes a processor, a memory, a communication interface, and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs are configured to be executed by the processor according to the following instructions:

[0138] Obtain basic filter parameter information;

[0139] Construct a three-dimensional Kalman filter coefficient table based on the basic filter parameter information;

[0140] Obtain observation information of the current frame detection box for the current target, as well as historical distance data and historical velocity data of the current target;

[0141] The cumulative deviation of appearance features is determined based on the observation information of the detection box in the current frame;

[0142] Based on the historical distance data and historical speed data, determine the cumulative distance deviation and cumulative speed deviation;

[0143] The stability level of the current target detection is determined based on the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the velocity.

[0144] Based on the stability level and the three-dimensional Kalman filter coefficient table, determine the target Kalman filter strength value of the current target at the current moment;

[0145] The current target's state is updated based on the Kalman filter strength value.

[0146] As can be seen, in this embodiment, the electronic device first acquires basic filtering parameter information, then constructs a three-dimensional Kalman filter coefficient table based on the basic filtering parameter information, then acquires the observation information of the current frame detection box for the current target, and acquires the historical distance data and historical velocity data of the current target. Further, based on the observation information of the current frame detection box, it determines the cumulative deviation of appearance features. Further still, based on the historical distance data and historical velocity data, it determines the cumulative distance deviation and cumulative velocity deviation. Then, based on the cumulative deviation of appearance features, cumulative distance deviation, and cumulative velocity deviation, it determines the stability level of the current target detection. Next, based on the stability level and the three-dimensional Kalman filter coefficient table, it determines the target Kalman filter strength value of the current target at the current moment. Finally, it updates the state of the current target based on the Kalman filter strength value. By converting the effective information of the target into gain coefficients for the Kalman coefficient table, and converting the Kalman filter into a coefficient table with fixed coefficients, and by determining the three-dimensional index in real time, a lookup table method is used to replace each matrix inversion operation, which helps to improve computational efficiency.

[0147] The above primarily describes the solutions of the embodiments of this application from the perspective of the method execution process. It is understood that, in order to achieve the above functions, the electronic device includes corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments provided herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0148] This application embodiment can divide the electronic device into functional units according to the above method example. For example, each function can be divided into a separate functional unit, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional unit. It should be noted that the unit division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0149] When dividing each function into modules according to its corresponding function. Figure 8 A functional unit block diagram of a Kalman filter device is given, such as... Figure 8As shown, the Kalman filter device includes an acquisition unit 801, a construction unit 802, a determination unit 803, and an update unit 804; wherein,

[0150] The acquisition unit 801 is used to acquire basic filter parameter information;

[0151] The construction unit 802 is used to construct a three-dimensional Kalman filter coefficient table based on the basic filter parameter information.

[0152] The acquisition unit 801 is also used to acquire observation information of the current frame detection box for the current target, and to acquire historical distance data and historical speed data of the current target;

[0153] The determining unit 803 is used to determine the cumulative deviation of appearance features based on the observation information of the detection box in the current frame;

[0154] The determining unit 803 is further configured to determine the cumulative distance deviation and the cumulative speed deviation based on the historical distance data and the historical speed data;

[0155] The determining unit 803 is further configured to determine the stability level of the current target detection based on the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the velocity.

[0156] The determining unit 803 is further configured to determine the target Kalman filter strength value of the current target at the current moment based on the stability level and the three-dimensional Kalman filter coefficient table;

[0157] The update unit 804 is used to update the current target's state based on the Kalman filter intensity value.

[0158] As can be seen from the embodiments of this application, the Kalman filter device first acquires basic filtering parameter information, then constructs a three-dimensional Kalman filter coefficient table based on the basic filtering parameter information, then acquires the observation information of the current frame detection box for the current target, and acquires the historical distance data and historical velocity data of the current target. Further, based on the observation information of the current frame detection box, it determines the cumulative deviation of appearance features. Further still, based on the historical distance data and historical velocity data, it determines the cumulative distance deviation and cumulative velocity deviation. Then, based on the cumulative deviation of appearance features, cumulative distance deviation, and cumulative velocity deviation, it determines the stability level of the current target detection. Next, based on the stability level and the three-dimensional Kalman filter coefficient table, it determines the target Kalman filter intensity value of the current target at the current moment. Finally, it updates the state of the current target based on the Kalman filter intensity value. By converting the effective information of the target into gain coefficients for the Kalman coefficient table, and converting the Kalman filter into a coefficient table with fixed coefficients, and by determining the three-dimensional index in real time, a lookup table method is used to replace each matrix inversion operation, which helps to improve computational efficiency.

[0159] It should be noted that all relevant content of each step involved in the above method embodiments can be referenced from the functional description of the corresponding functional module, and will not be repeated here.

[0160] The electronic device provided in this embodiment is used to execute the Kalman filtering method described above, and therefore can achieve the same effect as the implementation method described above.

[0161] When using integrated units, the electronic device may include a processing module, a storage module, and a communication module. The processing module can be used to control and manage the actions of the electronic device; for example, it can support the electronic device in executing the steps performed by the aforementioned functional units. The storage module can support the electronic device in executing stored program code and data. The communication module can support communication between the electronic device and other devices.

[0162] The processing module can be a processor or a controller. It can implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of digital signal processing (DSP) and a microprocessor, etc. The storage module can be a memory. The communication module can specifically be a radio frequency circuit, a Bluetooth chip, a Wi-Fi chip, or other devices that interact with other electronic devices.

[0163] This application also provides a computer storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the methods described in the above method embodiments, wherein the computer includes an electronic device.

[0164] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of any of the methods described in the above method embodiments. The computer program product may be a software installation package, and the computer includes a control platform.

[0165] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0166] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0167] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0168] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0169] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0170] If the aforementioned integrated units are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0171] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage device, which may include: a flash drive, a read-only memory, a random access memory, a magnetic disk, or an optical disk, etc.

[0172] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A Kalman filtering method characterized by, include: Obtain basic filter parameter information; Construct a three-dimensional Kalman filter coefficient table based on the basic filter parameter information; Obtain observation information of the current frame detection box for the current target, as well as historical distance data and historical velocity data of the current target; The cumulative deviation of appearance features is determined based on the observation information of the detection box in the current frame; Based on the historical distance data and historical speed data, determine the cumulative distance deviation and cumulative speed deviation; The stability level of the current target detection is determined based on the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the velocity. Based on the stability level and the three-dimensional Kalman filter coefficient table, determine the target Kalman filter strength value of the current target at the current moment; The current target's state is updated based on the Kalman filter strength value; Based on the cumulative deviation of the appearance features, the cumulative deviation of distance, and the cumulative deviation of velocity, the stability level of the current target detection is determined, including: Based on the observation information, determine the occlusion attributes; Based on the occlusion attributes, determine the fusion weights of the cumulative deviation of the appearance features, the cumulative deviation of the distance, and the cumulative deviation of the velocity; The feature fluctuation fusion value is determined based on the fusion weight, the occlusion attribute, the cumulative deviation of appearance features, the cumulative deviation of distance, and the cumulative deviation of velocity. Based on the fusion value of the characteristic fluctuations, the stability level of the current target detection is determined; Based on the stability level and the three-dimensional Kalman filter coefficient table, determine the target Kalman filter strength value of the current target at the current time, including: Based on the stability level, determine the target Kalman filter strength level; Based on the observation information, determine the time index and target noise level. The target Kalman filter strength value of the current target at the current time is obtained by querying the three-dimensional Kalman filter coefficient table based on the target Kalman filter strength level, the time index, and the target noise level.

2. The method of claim 1, wherein, The basic filtering parameter information includes the current target tracking distance range, preset Kalman noise, and operating condition information; The construction of the three-dimensional Kalman filter coefficient table based on the basic filter parameter information includes: Based on the tracking distance range, multiple noise levels are set, and each noise level corresponds to a set of Kalman filter noise parameters; A two-dimensional Kalman parameter table is determined based on the preset Kalman noise and the Kalman filter noise parameters corresponding to each noise level. Based on the operating condition information, multiple Kalman filter strength levels are determined, and the filtering factor for each Kalman filter strength level is generated. Based on the filter ratio and the two-dimensional Kalman parameter table, determine the three-dimensional Kalman filter coefficient table.

3. The method of claim 2, wherein, The observation information includes the width ratio of the current frame width to the previous frame width and the height ratio of the current frame height to the previous frame height of the current frame detection box. The cumulative deviation of the appearance features includes the cumulative deviation of the width ratio and the cumulative deviation of the height ratio. Determining the cumulative deviation of the appearance features based on the observation information of the current frame detection box includes: Acquire historical observation information, which includes historical width ratio data and historical height ratio data of the detection box. The historical width ratio data includes multiple historical width ratios, and the historical height ratio data includes multiple historical height ratios. By fitting the historical width ratio data, a first straight line is obtained; By fitting the historical height ratio data, a second straight line is obtained; Determine the cumulative deviation of the current frame width ratio from the first straight line; Determine the cumulative deviation of the current frame height ratio from the height ratio of the second straight line.

4. The method of claim 2, wherein, The step of determining the cumulative distance deviation and cumulative speed deviation based on the historical distance data and historical speed data includes: By fitting the historical distance data, a first curve is obtained; By fitting the historical velocity data, a second curve is obtained; Determine the cumulative distance deviation of the current frame from the first curve; Determine the cumulative velocity deviation of the current frame rate from the second curve.

5. The method of claim 1, wherein, The occlusion attribute is negatively correlated with the fusion weight of the cumulative deviation of the appearance feature, and positively correlated with the fusion weight of the cumulative distance deviation and the cumulative speed deviation.

6. A Kalman filter device, characterized in that, The Kalman filter device includes an acquisition unit, a construction unit, a determination unit, and an update unit; wherein... The acquisition unit is used to acquire basic filter parameter information; The construction unit is used to construct a three-dimensional Kalman filter coefficient table based on the basic filter parameter information. The acquisition unit is also used to acquire observation information of the current frame detection box for the current target, and to acquire historical distance data and historical velocity data of the current target; The determining unit is used to determine the cumulative deviation of appearance features based on the observation information of the detection box in the current frame; The determining unit is further configured to determine the cumulative distance deviation and the cumulative speed deviation based on the historical distance data and the historical speed data; The determining unit is further configured to determine the stability level of the current target detection based on the cumulative deviation of appearance features, the cumulative deviation of distance, and the cumulative deviation of velocity; determining the stability level of the current target detection based on the cumulative deviation of appearance features, the cumulative deviation of distance, and the cumulative deviation of velocity includes: determining an occlusion attribute based on observation information; determining a fusion weight of the cumulative deviation of appearance features, the cumulative deviation of distance, and the cumulative deviation of velocity based on the occlusion attribute; determining a feature fluctuation fusion value based on the fusion weight, the occlusion attribute, the cumulative deviation of appearance features, the cumulative deviation of distance, and the cumulative deviation of velocity; and determining the stability level of the current target detection based on the feature fluctuation fusion value. The determining unit is further configured to determine the target Kalman filter strength value of the current target at the current time based on the stability level and the three-dimensional Kalman filter coefficient table; determining the target Kalman filter strength value of the current target at the current time based on the stability level and the three-dimensional Kalman filter coefficient table includes: determining the target Kalman filter strength level based on the stability level; determining the time index and target noise level based on the observation information; and querying the three-dimensional Kalman filter coefficient table based on the target Kalman filter strength level, the time index, and the target noise level to obtain the target Kalman filter strength value of the current target at the current time. The update unit is used to update the state of the current target based on the Kalman filter intensity value.

7. An electronic device, characterized in that, It includes a processor and a memory, the memory being used to store one or more programs and configured to be executed by the processor, the programs including instructions for performing the steps of the method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, A computer program for storing electronic data interchange is provided, wherein the computer program causes a computer to perform the method as described in any one of claims 1-5.

Images