CN114048333A - Multi-source fusion voice interactive indoor positioning method, terminal and storage medium - Google Patents
Multi-source fusion voice interactive indoor positioning method, terminal and storage medium Download PDFInfo
- Publication number
- CN114048333A CN114048333A CN202111306697.1A CN202111306697A CN114048333A CN 114048333 A CN114048333 A CN 114048333A CN 202111306697 A CN202111306697 A CN 202111306697A CN 114048333 A CN114048333 A CN 114048333A
- Authority
- CN
- China
- Prior art keywords
- voice
- position description
- indoor
- probability
- nearby
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000004927 fusion Effects 0.000 title claims abstract description 58
- 230000002452 interceptive effect Effects 0.000 title claims abstract description 55
- 238000000034 method Methods 0.000 title claims abstract description 55
- 230000007704 transition Effects 0.000 claims abstract description 26
- 230000006870 function Effects 0.000 claims description 72
- 238000005315 distribution function Methods 0.000 claims description 9
- 238000005192 partition Methods 0.000 claims description 7
- 230000011218 segmentation Effects 0.000 claims description 5
- 238000000638 solvent extraction Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 7
- 238000004590 computer program Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 3
- 230000019771 cognition Effects 0.000 description 3
- 230000001149 cognitive effect Effects 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 230000004807 localization Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 241000233805 Phoenix Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000003930 cognitive ability Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000452 restraining effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/38—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/387—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using geographical or spatial information, e.g. location
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
- G06F40/35—Discourse or dialogue representation
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Physics & Mathematics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Library & Information Science (AREA)
- Signal Processing (AREA)
- Navigation (AREA)
Abstract
The invention discloses a multi-source fusion voice interactive indoor positioning method, a terminal and a storage medium, wherein the method comprises the following steps: acquiring voice position description information, and extracting a reference object in the voice position description information and a corresponding nearby spatial relationship; establishing an indoor position description classification frame according to the indoor position description rule and the number of nearby spatial relations; modeling the nearby spatial relationship, and positioning according to the indoor position description classification frame to obtain position description positioning information of the target object; and establishing a characteristic function of each observation value, and calculating the next moment position of the target object based on the state transition probability of the pedestrian dead reckoning so as to perform multi-source fusion positioning according to the hidden Markov model. The invention establishes an indoor position description classification frame according to the acquired voice position description information, models the nearby spatial relationship, and can realize a low-cost high-precision voice interactive indoor positioning mode on the basis of the existing hardware equipment.
Description
Technical Field
The invention relates to the field of terminal application, in particular to a multi-source fusion voice interactive indoor positioning method, a terminal and a storage medium.
Background
With the development of science and technology, in recent years, with the rapid development of mobile sensors and wireless networks, indoor positioning technology becomes a common research hotspot in multiple fields such as mobile internet, internet of things, mobile computing and location services, smart home, smart buildings and the like.
There are a variety of indoor positioning technologies, including: WiFi, Bluetooth, geomagnetism, Ultra Wideband (UWB), RFID, inertial navigation, and the like. The UWB positioning technology can reach the decimeter level and even higher positioning precision; WiFi, Bluetooth, geomagnetism and other positioning technologies can reach meter-level positioning accuracy; inertial navigation requires no additional equipment, but requires the known navigation starting point position, and therefore cannot be used alone. Currently, internet macros including hundredths, Tencent, Ali, apple, etc. are all in the field of indoor positioning in corner, but there is still no large-scale commercialized and reliable indoor positioning service. The main reason for this is that these positioning technologies (bluetooth, UWB, RFID, etc.) have a strong dependence on the infrastructure, resulting in high positioning costs. WiFi, geomagnetism and PDR positioning methods are influenced by indoor complex topological structures, and based on the single positioning source, the realization of indoor positioning meeting various requirements still has great challenges, which are reflected in low precision, poor stability and the like.
The mobile phone sensor can sense the surrounding environment and user behaviors, and a method for improving the positioning accuracy or stability through semantic sensing is concerned by more and more people. With the mining of the user spatial cognitive abilities, indoor positioning technology is gradually developing from perception to cognition. Under the background of the artificial intelligence era, the requirements of users on indoor positioning are not only precision, but also better fusion with intelligent terminal equipment and intelligent service. Therefore, on the premise of considering positioning accuracy, the intelligent low-cost indoor positioning method is developed by combining space cognition, and has important practical significance.
Therefore, the prior art has yet to be improved.
Disclosure of Invention
The invention provides a multi-source fusion voice interactive indoor positioning method, a terminal and a storage medium, aiming at solving the technical problems of high cost and low precision of the existing indoor positioning mode.
The technical scheme adopted by the invention for solving the technical problem is as follows:
in a first aspect, the present invention provides a multi-source fusion voice interactive indoor positioning method, which comprises the following steps:
acquiring voice position description information, and extracting a reference object in the voice position description information and a corresponding nearby spatial relationship;
establishing an indoor position description classification frame according to an indoor position description rule and the number of the nearby spatial relations;
modeling the nearby spatial relationship, and positioning according to the indoor position description classification frame to obtain position description positioning information of the target object;
and establishing a characteristic function of each observation value, and calculating the position of the target object at the next moment based on the state transition probability of the pedestrian dead reckoning so as to perform multi-source fusion positioning according to the hidden Markov model.
In one implementation, the voice location description information includes: the reference object, the nearby spatial relationship, and the target object;
the target object is a position to be positioned.
In one implementation, the acquiring the voice position description information and extracting the reference object and the corresponding nearby spatial relationship thereof in the voice position description information includes:
acquiring the voice position description information;
inputting the voice position description information into a voice recognition module, and converting the voice position description information into corresponding text information through the voice recognition module;
and performing word segmentation processing on the text information through a word bag model, and extracting a reference object in the text information and a corresponding nearby spatial relationship.
In one implementation, the establishing an indoor location description classification framework according to an indoor location description rule and the number of nearby spatial relationships includes:
determining the number of the nearby spatial relationships in the voice position description information according to the indoor position description rule;
and partitioning according to the number of the nearby spatial relations, and establishing the indoor position description classification frame according to the partition corresponding to each nearby spatial relation.
In one implementation, the modeling the nearby spatial relationship and locating according to the indoor location description classification frame to obtain location description location information of the target object includes:
constructing a probability density function according to the nearby spatial relationship;
judging whether the membership function of the nearby spatial relationship meets a preset probability constraint condition or not;
and if the membership function of the nearby spatial relationship meets the preset probability constraint condition, calculating the position description positioning information of the target object according to the indoor position description classification frame and a joint probability formula.
In one implementation, the determining whether the membership function of the nearby spatial relationship satisfies a preset probability constraint condition includes:
judging whether the probability of the membership function at the boundary of the reference object is a first preset value or not;
if the probability of the membership function at the reference object boundary is the first preset value, judging whether the probability of the membership function in a non-reference object nearby area is a second preset value;
if the probability of the membership function in the area near the non-reference object is the second preset value, judging whether the continuity of the membership function tends to a third preset value;
and if the continuity of the membership function tends to the third preset value, judging that the membership function meets the preset probability constraint condition.
In one implementation, the establishing a feature function of each observation value and calculating a next-moment position of the target object based on a state transition probability of pedestrian dead reckoning to perform multi-source fusion positioning according to a hidden markov model includes:
establishing a characteristic function of each observation value to obtain multi-source observation data;
fusing the multi-source observation data through a conditional random model to obtain an observation probability;
determining the state transition probability of the pedestrian dead reckoning according to a positive-high Gaussian distribution function;
and calculating the position of the target object at the next moment according to the observation probability and the state transition probability of the pedestrian dead reckoning so as to perform multi-source fusion positioning according to the hidden Markov model.
In one implementation, the determining the state transition probability of the pedestrian dead reckoning according to a positive Gaussian distribution function includes:
determining observation distances and observation angles corresponding to the observation values according to a positive-too-high Gaussian distribution function;
and calculating the state transition probability of the pedestrian dead reckoning according to the observation distance and the observation angle.
In a second aspect, the present invention provides a terminal, comprising: the multi-source fusion voice interactive indoor positioning method comprises a processor and a memory, wherein the memory stores a multi-source fusion voice interactive indoor positioning program, and the multi-source fusion voice interactive indoor positioning program is used for realizing the multi-source fusion voice interactive indoor positioning method in the first aspect when being executed by the processor.
In a third aspect, the present invention provides a storage medium, where a multi-source fusion voice interactive indoor positioning program is stored, and the multi-source fusion voice interactive indoor positioning program is used to implement the multi-source fusion voice interactive indoor positioning method according to the first aspect when being executed by a processor.
The invention adopts the technical scheme and has the following effects:
the method and the device acquire the voice position description information in a voice interaction mode, can establish an indoor position description classification frame according to the voice position description information, and can perform positioning according to the indoor position description classification frame to obtain the position description positioning information of the target object; moreover, the invention can calculate the next moment position of the target object based on the state transition probability of the pedestrian dead reckoning by establishing the characteristic function of each observation value, thereby realizing multi-source fusion positioning and realizing a low-cost high-precision voice interactive indoor positioning mode on the basis of the existing hardware equipment.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the structures shown in the drawings without creative efforts.
FIG. 1 is a flow chart of a multi-source fusion voice interactive indoor positioning method in one implementation of the present invention.
FIG. 2 is a cognitive schematic of qualitative distance in one implementation of the invention.
FIG. 3 is a schematic illustration of a partition of a reference object in one implementation of the invention.
FIG. 4 is a schematic diagram of a multi-source data fusion model in one implementation of the invention.
Fig. 5 is a functional schematic of a terminal in one implementation of the invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer and clearer, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Exemplary method
As shown in fig. 1, an embodiment of the present invention provides a multi-source fusion voice interactive indoor positioning method, where the multi-source fusion voice interactive indoor positioning method includes the following steps:
step S100, acquiring voice position description information, and extracting a reference object in the voice position description information and a corresponding nearby spatial relationship.
In this embodiment, the multi-source fusion voice interactive indoor positioning method is applied to a terminal, where the terminal includes but is not limited to: computers, mobile terminals, and the like.
In this embodiment, the multi-source fusion voice interactive indoor positioning method is a multi-source fusion voice interactive indoor positioning method based on near spatial relationship enhancement, and mainly solves the problems that the traditional indoor positioning method is single in interactive mode, poor in positioning accuracy and the like due to the fact that a single signal source is used; by the multi-source fusion voice interactive indoor positioning method, a low-cost high-precision voice interactive indoor positioning mode can be realized on the basis of the existing hardware equipment.
In this embodiment, the principle of the multi-source fusion voice interactive indoor positioning method lies in: firstly, a conversion process from voice position description to character position description is completed through a voice recognition module, and then a reference object in the voice position description and a related 'nearby' spatial relationship are extracted through word segmentation; then, an indoor position description classification frame is established by combining the indoor position description habit and the 'nearby' spatial relationship; carrying out uncertainty modeling and positioning by utilizing an indoor 'nearby' spatial relationship; and finally, establishing a characteristic function of each observation value based on a hidden Markov model to realize indoor accurate positioning.
Specifically, when the multi-source fusion voice interactive indoor positioning method is implemented, voice position description information input by a target object (namely, a testee needing to be positioned) needs to be acquired through a voice acquisition device, and then the voice position description information is input into a voice module so as to be converted into character position description information through the voice module; the voice location description information needs to include descriptors of spatial location relationships such as "nearby", for example: "i am near mcdonald's.
In this embodiment, in the process of converting the voice position description information into the text position description information, the conversion process from voice to text may be completed through a voice recognition module such as a scientific news flyer or a hectometer; then, the word bag model is used for segmenting the converted characters to identify a required reference object and identify a near spatial relation related to the reference object, and finally, the reference object is associated with the near spatial relation.
Further, since the target object in the room needs to be located, the voice position description information in the room needs to include: a Reference Object, a "Spatial relationship" related to the Reference Object, and a Target Object. Wherein the reference object is a feature having a name in a room, such as: mcdonald green in the market; the target object is a description position, namely: the location to be located. The spatial relationship includes: topological relations, orientation relations and distance relations (quantitative distances and qualitative distances).
In this embodiment, the position clue transmitted by the topological relation is rough, and the distance relation and the orientation relation can indirectly reflect the topological relation; thus, topological relationships are not generally used for position descriptive positioning.
Because of the lack of absolute reference indoors, people use more relative orientation relationships for position description, namely: "front, back, left, right", or "front, back, left, right, left front, right back, left back", etc. The distance relationship includes: quantitative distances (e.g., 50m) and qualitative distances (e.g., nearby). Therefore, in an indoor location description scene, the application frequency of the "nearby" spatial relationships in the spatial relationship categories is the highest; the present embodiment takes "vicinity" as a representative word of the spatial relationship of the "vicinity".
That is, in an implementation manner of this embodiment, step S100 specifically includes the following steps:
step S110, acquiring the voice position description information;
step S120, inputting the voice position description information into a voice recognition module, and converting the voice position description information into corresponding text information through the voice recognition module;
step S130, performing word segmentation processing on the text information through a word bag model, and extracting a reference object in the text information and a corresponding nearby spatial relationship.
In the embodiment, the voice position description information is converted into the text position description information, so that the voice interaction function can be integrated into the indoor positioning function; moreover, the word segmentation processing is carried out on the character information, so that the reference object and the corresponding 'nearby' spatial relationship can be extracted.
As shown in fig. 1, in an implementation manner of the embodiment of the present invention, the multi-source fusion voice interactive indoor positioning method further includes the following steps:
and step S200, establishing an indoor position description classification frame according to the indoor position description rule and the number of the nearby spatial relations.
In this embodiment, after the reference objects and the corresponding "nearby" spatial relationship are extracted, the number of the reference objects in the speech position description information may be determined according to the indoor position description rule. The indoor position description rule is the description habit of people for the indoor position; and then, determining the number of the spatial relations of the 'vicinity' of the reference objects according to the number of the reference objects.
Further, from the description convention analysis, the number of the reference objects in the complete position description at least includes one, and is not more than three. Here, the unary represents the number of reference objects as one, and the position description is divided into an unary reference object position description, a binary reference object position description, and a ternary reference object position description according to the number of reference objects based on the "nearby" spatial relationship. For example: the unary reference object location is described as "i am near mcdonald"; the binary reference object location is described as "i am near mcdonald, near zuka"; the ternary reference object location is described as "i am in mcdonald (nearby), zakayi (nearby), and the old phoenix nearby".
In cognitive analysis, the spatial relationship of the 'vicinity' is the qualitative expression of people on the space distance cognition, and in order to facilitate calculation, the conversion from the qualitative to the quantitative needs to be completed. Human knowledge of the spatial relationship of "vicinity" is somewhat inversely proportional to distance, as shown in fig. 2, and within a certain distance range (d < d0), there is no change in the knowledge of "vicinity"; beyond a certain distance (d > d0), knowledge of "nearby" becomes weaker and weaker as the distance increases.
Partitioning the space where the target object is located according to the number of the 'nearby' spatial relations, and further establishing the indoor position description classification frame according to the partitions corresponding to the 'nearby' spatial relations; as shown in fig. 3, when partitioning is performed, the space may be partitioned according to the point sets, each partition is composed of the point set closest to the point, and has a good potential in expressing the spatial proximity relationship.
That is, in an implementation manner of this embodiment, the step S200 specifically includes the following steps:
step S210, determining the number of the nearby spatial relationships in the voice position description information according to the indoor position description rule;
and step S220, partitioning according to the number of the nearby spatial relations, and establishing the indoor position description classification frame according to the partition corresponding to each nearby spatial relation.
According to the method and the device, the number of the spatial relations of the vicinity in the voice position description information can be determined through the indoor position description rule, the partition is carried out according to the number of the spatial relations of the vicinity, an indoor position description classification frame is established according to the cognitive distance, and the target object is positioned on the basis of the classification frame.
As shown in fig. 1, in an implementation manner of the embodiment of the present invention, the multi-source fusion voice interactive indoor positioning method further includes the following steps:
and step S300, modeling the nearby spatial relationship, and positioning according to the indoor position description classification frame to obtain position description positioning information of the target object.
In this embodiment, after the indoor location description classification framework is established, a probability density function can be established according to the nearby spatial relationship, and then whether the membership function of the nearby spatial relationship meets a preset probability constraint condition is judged; and if the membership function of the nearby spatial relationship meets the preset probability constraint condition, calculating the position description positioning information of the target object according to the indoor position description classification frame and a joint probability formula.
Further, in constructing the probability density function, the reference object may be set as a "face" object, that is, the reference object may be thought of as a face; based on the principle of Euclidean distance and 'stolen-area', the 'nearby' probability density function of 'point' (namely target object) is expanded. On the basis, whether the membership function of the nearby spatial relationship meets a preset probability constraint condition or not is judged according to a preset probability constraint condition, and under the condition that the preset probability constraint condition is met, the position description positioning information of the target object is calculated according to the indoor position description classification frame and a joint probability formula.
That is, in an implementation manner of this embodiment, the step S300 specifically includes the following steps:
step S310, a probability density function is constructed according to the nearby space relation;
step S320, judging whether the membership function of the nearby spatial relationship meets a preset probability constraint condition or not;
step S330, if the membership function of the nearby spatial relationship meets the preset probability constraint condition, calculating the position description positioning information of the target object according to the indoor position description classification frame and a joint probability formula.
In this embodiment, when the membership function satisfies the predetermined probability constraint condition, it may be determined whether the probability of the membership function at the reference object boundary is a first preset value, where the first preset value is 1; if the probability of the membership function at the reference object boundary is the first preset value, whether the probability of the membership function in a non-reference object nearby area is a second preset value is also required to be judged, wherein the second preset value is 0; if the probability of the membership function in the area near the non-reference object is the second preset value, judging whether the continuity of the membership function tends to a third preset value, wherein the third preset value is 0; and if the continuity of the membership function tends to the third preset value, judging that the membership function meets the preset probability constraint condition.
Based on the function level, the membership function of the spatial relationship near the 'surface' object needs to satisfy the following constraint.
Restraining one: the probability is 1 at the reference object boundary (see equation 1);
and (2) constraining: the probability is 0 in the "vicinity" area of the non-reference object (see formula 2, where d (Ri) represents the "vicinity" area of the reference object Ri);
and (3) constraining: continuity (see equation 3).
p(x，y，Ri)＝1 (1)
Further, according to the indoor position description classification framework, the position relation of the indoor position description of the 'nearby' spatial relation can be obtained based on a joint probability formula (see formula 4), and the position information of the reference object can be obtained; where n denotes the number of reference objects and p (ri) denotes the probability of a "nearby" spatial relationship with respect to a reference object.
That is, in an implementation manner of the present embodiment, in step S320, the following steps are specifically included:
step S321, judging whether the probability of the membership function at the reference object boundary is a first preset value;
step S322, if the probability of the membership function at the reference object boundary is the first preset value, judging whether the probability of the membership function in the area close to the non-reference object is a second preset value;
step S323, if the probability of the membership function in the area near the non-reference object is the second preset value, judging whether the continuity of the membership function tends to a third preset value;
step S324, if the continuity of the membership function tends to the third preset value, determining that the membership function satisfies the preset probability constraint condition.
In the embodiment, by constructing the probability density function, when the membership function of the "nearby" spatial position relation meets the preset probability constraint condition, the position description positioning information of the target object can be calculated according to the indoor position description classification frame and the joint probability formula.
As shown in fig. 1, in an implementation manner of the embodiment of the present invention, the multi-source fusion voice interactive indoor positioning method further includes the following steps:
and S400, establishing a characteristic function of each observation value, and calculating the next moment position of the target object based on the state transition probability of the pedestrian dead reckoning so as to perform multi-source fusion positioning according to the hidden Markov model.
In this embodiment, after the position description positioning information of the target object is obtained, a feature function of each observation value can be established to obtain multi-source observation data; then, fusing the multi-source observation data through a conditional random model to obtain an observation probability; and determining the state transition probability of the pedestrian dead reckoning according to a positive-high Gaussian distribution function; and finally, calculating the position of the target object at the next moment according to the observation probability and the state transition probability of the pedestrian dead reckoning so as to perform multi-source fusion positioning according to the hidden Markov model.
Specifically, in the present embodiment, the location-descriptive localization of the pedestrian (i.e., the descriptive localization of the target object) can be described by using a Hidden Markov (HMM) process; therefore, the present embodiment uses a hidden markov factor graph to model the "nearby" spatial relationship location description positioning process.
As shown in FIG. 4, V0，V1，V2……VmRepresents the indoor position of the pedestrian at the time of 0, 1, m;
k kinds of multi-source observation data representing m moments;
indicating m is at position VmThe observed probability of the kth observed data is obtained by calculation of a characteristic function of the observed data; represents the time p (V)m|Vm-1) To VmThe state transition probability of (2) is obtained from an observation result of the dead reckoning of the pedestrian.
That is, in an implementation manner of this embodiment, the step S400 specifically includes the following steps:
step S410, establishing a characteristic function of each observation value to obtain multi-source observation data;
step S420, fusing the multi-source observation data through a conditional random model to obtain an observation probability;
step S430, determining the state transition probability of the pedestrian dead reckoning according to a positive-high Gaussian distribution function;
step S440, calculating the position of the target object at the next moment according to the observation probability and the state transition probability of the pedestrian dead reckoning, so as to perform multi-source fusion positioning according to the hidden Markov model.
In this embodiment, building a hidden markov factor graph model requires parameterized expression of observation factors, i.e., building a feature function of each observation value.
The observation factors in this embodiment include "nearby" spatial relationship description positioning, WiFi positioning, and geomagnetic positioning. The positioning result described by the space relation of the 'nearby' can be obtained by the position description classification result based on the normalization of the joint probability (formula 4), and is expressed as
WiFi positioning is realized based on a fingerprint algorithm, Euclidean distances between signal intensity received by the current position of a pedestrian and signal intensity in a sub-fingerprint database based on near spatial relation constraint are calculated, then the Euclidean distances are subjected to reciprocal normalization to obtain the probability that the pedestrian is located at the current position, and a characteristic function is expressed as
The geomagnetic positioning compares the Euclidean distance of the geomagnetic sequence (DTW) with the WiFi positioning, and the characteristic function is expressed as
On the basis of a single-source observation data characteristic function, fusing multi-source observation data by adopting a conditional random field model:
According to Bayesian theorem P (D)m|Vm)＝P(Vm|Dm)·P(Dm)/P(Vm) The following can be obtained:
the indoor pedestrian position description positioning process can be expressed as that the position of the next moment is calculated according to the state transition probability on the basis of the characteristic function (namely, observation probability) of the multi-source observation data fusion model and the state transition probability of the pedestrian dead reckoning. The observation result of the pedestrian dead reckoning comprises an observation distance and an observation angle, and can be described by positive and high Gaussian distribution, the two are mutually independent, and the state transition probability is as follows:
wherein sigmadAnd σeRespectively, the standard deviation of the distance and direction. Thus, the multi-source fusion localization model of indoor pedestrian location description can be expressed as:
P(Vm)＝P(Vm-1)·P(Vm|Vm-1)·P(Vm|Dm) (7)
initial position (V)0) According to the acquisition of a plurality of groups of observation data, the probability P (V) of the observation data is calculated based on the characteristic function0)。
That is, in an implementation manner of this embodiment, the step S430 specifically includes the following steps:
step S431, determining observation distances and observation angles corresponding to the observation values according to a positive-high Gaussian distribution function;
step S432, calculating the state transition probability of the pedestrian dead reckoning according to the observation distance and the observation angle.
In the embodiment, the voice position description information is acquired in a voice interaction mode, an indoor position description classification frame can be established according to the voice position description information, and positioning is performed according to the indoor position description classification frame, so that the position description positioning information of the target object can be obtained; moreover, the invention can calculate the next moment position of the target object based on the state transition probability of the pedestrian dead reckoning by establishing the characteristic function of each observation value, thereby realizing multi-source fusion positioning and realizing a low-cost high-precision voice interactive indoor positioning mode on the basis of the existing hardware equipment.
Exemplary device
Based on the above embodiments, the present invention further provides a terminal, and a schematic block diagram thereof may be as shown in fig. 5.
The terminal includes: the system comprises a processor, a memory, an interface, a display screen and a communication module which are connected through a system bus; wherein the processor of the terminal is configured to provide computing and control capabilities; the memory of the terminal comprises a storage medium and an internal memory; the storage medium stores an operating system and a computer program; the internal memory provides an environment for the operation of an operating system and a computer program in the storage medium; the interface is used for connecting external terminal equipment, such as mobile terminals and computers; the display screen is used for displaying corresponding multi-source fusion voice interactive indoor positioning information; the communication module is used for communicating with a cloud server or a mobile terminal.
The computer program is used for realizing a multi-source fusion voice interactive indoor positioning method when being executed by a processor.
It will be understood by those skilled in the art that the block diagram of fig. 5 is a block diagram of only a portion of the structure associated with the inventive arrangements and is not intended to limit the terminals to which the inventive arrangements may be applied, and that a particular terminal may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a terminal is provided, which includes: the multi-source fusion voice interactive indoor positioning method comprises a processor and a memory, wherein the memory stores a multi-source fusion voice interactive indoor positioning program, and the multi-source fusion voice interactive indoor positioning program is used for realizing the multi-source fusion voice interactive indoor positioning method when being executed by the processor.
In one embodiment, a storage medium is provided, wherein the storage medium stores a multi-source fusion voice interactive indoor positioning program, and the multi-source fusion voice interactive indoor positioning program is used for realizing the multi-source fusion voice interactive indoor positioning method when being executed by a processor.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware related to instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, the computer program can include the processes of the embodiments of the methods described above. Any reference to memory, storage, databases, or other media used in embodiments provided herein may include non-volatile and/or volatile memory.
In summary, the present invention provides a multi-source fusion voice interactive indoor positioning method, a terminal and a storage medium, wherein the method comprises: acquiring voice position description information, and extracting a reference object in the voice position description information and a corresponding nearby spatial relationship; establishing an indoor position description classification frame according to the indoor position description rule and the number of nearby spatial relations; modeling the nearby spatial relationship, and positioning according to the indoor position description classification frame to obtain position description positioning information of the target object; and establishing a characteristic function of each observation value, and calculating the next moment position of the target object based on the state transition probability of the pedestrian dead reckoning so as to perform multi-source fusion positioning according to the hidden Markov model. The invention establishes an indoor position description classification frame according to the acquired voice position description information, models the nearby spatial relationship, and can realize a low-cost high-precision voice interactive indoor positioning mode on the basis of the existing hardware equipment.
It is to be understood that the invention is not limited to the examples described above, but that modifications and variations may be effected thereto by those of ordinary skill in the art in light of the foregoing description, and that all such modifications and variations are intended to be within the scope of the invention as defined by the appended claims.
Claims (10)
1. A multi-source fusion voice interactive indoor positioning method is characterized by comprising the following steps:
acquiring voice position description information, and extracting a reference object in the voice position description information and a corresponding nearby spatial relationship;
establishing an indoor position description classification frame according to an indoor position description rule and the number of the nearby spatial relations;
modeling the nearby spatial relationship, and positioning according to the indoor position description classification frame to obtain position description positioning information of the target object;
and establishing a characteristic function of each observation value, and calculating the position of the target object at the next moment based on the state transition probability of the pedestrian dead reckoning so as to perform multi-source fusion positioning according to the hidden Markov model.
2. The multi-source fused voice interactive indoor positioning method according to claim 1, wherein the voice location description information comprises: the reference object, the nearby spatial relationship, and the target object;
the target object is a position to be positioned.
3. The multi-source fusion voice interactive indoor positioning method according to claim 1, wherein the obtaining of the voice position description information and the extracting of the reference object in the voice position description information and the corresponding nearby spatial relationship thereof comprise:
acquiring the voice position description information;
inputting the voice position description information into a voice recognition module, and converting the voice position description information into corresponding text information through the voice recognition module;
and performing word segmentation processing on the text information through a word bag model, and extracting a reference object in the text information and a corresponding nearby spatial relationship.
4. The multi-source fusion voice interactive indoor positioning method according to claim 1, wherein the establishing an indoor location description classification framework according to indoor location description rules and the number of the nearby spatial relationships comprises:
determining the number of the nearby spatial relationships in the voice position description information according to the indoor position description rule;
and partitioning according to the number of the nearby spatial relations, and establishing the indoor position description classification frame according to the partition corresponding to each nearby spatial relation.
5. The multi-source fusion voice interactive indoor positioning method according to claim 1, wherein the modeling the nearby spatial relationship and positioning according to the indoor location description classification framework to obtain location description positioning information of a target object comprises:
constructing a probability density function according to the nearby spatial relationship;
judging whether the membership function of the nearby spatial relationship meets a preset probability constraint condition or not;
and if the membership function of the nearby spatial relationship meets the preset probability constraint condition, calculating the position description positioning information of the target object according to the indoor position description classification frame and a joint probability formula.
6. The multi-source fusion voice interactive indoor positioning method according to claim 5, wherein the determining whether the membership function of the nearby spatial relationship satisfies a preset probability constraint condition comprises:
judging whether the probability of the membership function at the boundary of the reference object is a first preset value or not;
if the probability of the membership function at the reference object boundary is the first preset value, judging whether the probability of the membership function in a non-reference object nearby area is a second preset value;
if the probability of the membership function in the area near the non-reference object is the second preset value, judging whether the continuity of the membership function tends to a third preset value;
and if the continuity of the membership function tends to the third preset value, judging that the membership function meets the preset probability constraint condition.
7. The multi-source fusion voice interactive indoor positioning method according to claim 1, wherein the establishing a feature function of each observation value and calculating a next moment position of the target object based on a state transition probability of pedestrian dead reckoning for multi-source fusion positioning according to a hidden markov model comprises:
establishing a characteristic function of each observation value to obtain multi-source observation data;
fusing the multi-source observation data through a conditional random model to obtain an observation probability;
determining the state transition probability of the pedestrian dead reckoning according to a positive-high Gaussian distribution function;
and calculating the position of the target object at the next moment according to the observation probability and the state transition probability of the pedestrian dead reckoning so as to perform multi-source fusion positioning according to the hidden Markov model.
8. The multi-source fused speech interactive indoor positioning method of claim 7, wherein the determining the state transition probability of the pedestrian dead reckoning according to the positive Gaussian distribution function comprises:
determining observation distances and observation angles corresponding to the observation values according to a positive-too-high Gaussian distribution function;
and calculating the state transition probability of the pedestrian dead reckoning according to the observation distance and the observation angle.
9. A terminal, comprising: a processor and a memory, the memory storing a multi-source fused voice interactive indoor positioning program, the multi-source fused voice interactive indoor positioning program, when executed by the processor, being configured to implement the multi-source fused voice interactive indoor positioning method of any one of claims 1-8.
10. A storage medium storing a multi-source fused voice interactive indoor positioning program, wherein the multi-source fused voice interactive indoor positioning program is used for implementing the multi-source fused voice interactive indoor positioning method according to any one of claims 1 to 8 when being executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111306697.1A CN114048333B (en) | 2021-11-05 | 2021-11-05 | Multisource fusion voice interactive indoor positioning method, terminal and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111306697.1A CN114048333B (en) | 2021-11-05 | 2021-11-05 | Multisource fusion voice interactive indoor positioning method, terminal and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114048333A true CN114048333A (en) | 2022-02-15 |
| CN114048333B CN114048333B (en) | 2024-06-04 |
Family
ID=80207393
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111306697.1A Active CN114048333B (en) | 2021-11-05 | 2021-11-05 | Multisource fusion voice interactive indoor positioning method, terminal and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114048333B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116562234A (en) * | 2023-03-30 | 2023-08-08 | 深圳市规划和自然资源数据管理中心 | Multi-source data fusion voice indoor positioning method and related equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100217593A1 (en) * | 2009-02-05 | 2010-08-26 | Seiko Epson Corporation | Program for creating Hidden Markov Model, information storage medium, system for creating Hidden Markov Model, speech recognition system, and method of speech recognition |
| CN107635204A (en) * | 2017-09-27 | 2018-01-26 | 深圳大学 | A kind of indoor fusion and positioning method and device of motor behavior auxiliary, storage medium |
| CN109087648A (en) * | 2018-08-21 | 2018-12-25 | 平安科技（深圳）有限公司 | Sales counter voice monitoring method, device, computer equipment and storage medium |
| CN109769296A (en) * | 2017-11-09 | 2019-05-17 | 深圳市交投科技有限公司 | A kind of indoor orientation method merging multi-source data |
| CN110956955A (en) * | 2019-12-10 | 2020-04-03 | 苏州思必驰信息科技有限公司 | Voice interaction method and device |
-
2021
- 2021-11-05 CN CN202111306697.1A patent/CN114048333B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100217593A1 (en) * | 2009-02-05 | 2010-08-26 | Seiko Epson Corporation | Program for creating Hidden Markov Model, information storage medium, system for creating Hidden Markov Model, speech recognition system, and method of speech recognition |
| CN107635204A (en) * | 2017-09-27 | 2018-01-26 | 深圳大学 | A kind of indoor fusion and positioning method and device of motor behavior auxiliary, storage medium |
| CN109769296A (en) * | 2017-11-09 | 2019-05-17 | 深圳市交投科技有限公司 | A kind of indoor orientation method merging multi-source data |
| CN109087648A (en) * | 2018-08-21 | 2018-12-25 | 平安科技（深圳）有限公司 | Sales counter voice monitoring method, device, computer equipment and storage medium |
| CN110956955A (en) * | 2019-12-10 | 2020-04-03 | 苏州思必驰信息科技有限公司 | Voice interaction method and device |
Non-Patent Citations (2)
| Title |
|---|
| 王彦坤;樊红;王伟玺;李晓明;张星;: "地标空间方向的位置描述定位模型", 测绘科学, no. 09, 16 September 2020 (2020-09-16) * |
| 陈逸灵;程艳芬;陈先桥;王红霞;李超: "PAD三维情感空间中的语音情感识别", 哈尔滨工业大学学报, vol. 50, no. 011, 31 December 2018 (2018-12-31) * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116562234A (en) * | 2023-03-30 | 2023-08-08 | 深圳市规划和自然资源数据管理中心 | Multi-source data fusion voice indoor positioning method and related equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114048333B (en) | 2024-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8155877B2 (en) | Location-to-landmark | |
| US20190147360A1 (en) | Learned model provision method, and learned model provision device | |
| WO2020098252A1 (en) | Route planning method, server, and computer storage medium | |
| CN110781413B (en) | Method and device for determining interest points, storage medium and electronic equipment | |
| US20230266470A1 (en) | Robot relocalization method and apparatus, and storage medium and electronic device | |
| US20230041943A1 (en) | Method for automatically producing map data, and related apparatus | |
| CN103389979A (en) | System, device and method for recommending classification lexicon in input method | |
| KR20210098880A (en) | Voice processing method, apparatus, device and storage medium for vehicle-mounted device | |
| US11443116B2 (en) | Electronic apparatus and control method thereof | |
| CN113947147A (en) | Training method and positioning method of target map model and related devices | |
| CN114111813B (en) | High-precision map element updating method and device, electronic equipment and storage medium | |
| CN116518960A (en) | Road network updating method, device, electronic equipment and storage medium | |
| CN114048333B (en) | Multisource fusion voice interactive indoor positioning method, terminal and storage medium | |
| Nowakowski et al. | Topological localization using Wi-Fi and vision merged into FABMAP framework | |
| CN113766633A (en) | Data processing method, data processing device, electronic equipment and storage medium | |
| CN112488157A (en) | Dialog state tracking method and device, electronic equipment and storage medium | |
| CN111653268A (en) | Man-machine voice interaction system and method for shopping cabin | |
| CN114416941B (en) | Knowledge graph-fused dialogue knowledge point determination model generation method and device | |
| CN114139069A (en) | Indoor positioning method and system based on voice interaction and electronic equipment | |
| CN112561956B (en) | Video target tracking method and device, electronic equipment and storage medium | |
| CN115130456A (en) | Sentence parsing and matching model training method, device, equipment and storage medium | |
| CN114492573A (en) | Map matching method and device based on relaxation iteration semantic features | |
| CN113360791A (en) | Interest point query method and device of electronic map, road side equipment and vehicle | |
| CN111261165A (en) | Station name identification method, device, equipment and storage medium | |
| CN111797183A (en) | Method and device for mining road attribute of information point and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |