CN114872056A

CN114872056A - House map generation method and device, cleaning assembly and cleaning equipment

Info

Publication number: CN114872056A
Application number: CN202210420696.8A
Authority: CN
Inventors: 张海湘; 柯南海; 周技锋; 林进; 孙涛
Original assignee: Midea Robozone Technology Co Ltd
Current assignee: Midea Robozone Technology Co Ltd
Priority date: 2022-04-21
Filing date: 2022-04-21
Publication date: 2022-08-09

Abstract

The invention provides a house map generation method and device, a cleaning assembly and cleaning equipment. The house map generation method comprises the following steps: acquiring an original map and a depth image of a target house; correcting an original map according to a preset segmentation algorithm; determining feature information and room attributes of a plurality of rooms of a target house according to the depth image; updating the corrected house map according to the feature information and the room attribute to obtain a target map; wherein the feature information corresponds to the room attribute. The house map generation method corrects the original house map through a preset segmentation algorithm, determines feature information and room attributes of each room according to house depth data, and updates the corrected house map. In this way, the accuracy and integrity of the generated house map are ensured, thereby ensuring the accuracy of the work control of the cleaning assembly.

Description

House map generation method and device, cleaning assembly and cleaning equipment

Technical Field

The invention relates to the technical field of robots, in particular to a house map generation method and device, a cleaning assembly and cleaning equipment.

Background

Among the prior art, when the clean subassembly is through laser radar data structure house map, because the height of clean subassembly is shorter, and exists a plurality of shelters in the house, lead to laser radar data can't detect house edge wall body data to unable accurate indoor map of restoreing, make the house map of final structure incomplete, lead to the subsequent work control of clean subassembly accurate inadequately, influence clean effect.

Disclosure of Invention

The present invention is directed to solving at least one of the problems of the prior art or the related art.

To this end, a first aspect of the present invention is to propose a house map generation method.

A second aspect of the present invention is to provide a house map generating apparatus.

A third aspect of the invention is directed to a cleaning assembly.

A fourth aspect of the invention is directed to a cleaning assembly.

A fifth aspect of the present invention is to provide a cleaning apparatus.

A sixth aspect of the invention is directed to a readable storage medium.

In view of the above, according to an aspect of the present invention, a house map generation method is provided, including: acquiring an original map and a depth image of a target house; correcting the original map according to a preset segmentation algorithm; determining feature information and room attributes of a plurality of rooms of a target house according to the depth image; updating the corrected house map according to the feature information and the room attribute to obtain a target map; wherein the feature information corresponds to the room attribute.

The execution subject of the technical scheme of the house map generation method provided by the invention can be a house map generation device, and can also be determined according to actual use requirements, which is not specifically limited herein. In order to more clearly describe the house map generation method provided by the present invention, the following description will be made with an execution subject of the house map generation method as a house map generation apparatus.

According to the house map generation method provided by the invention, when the house map is generated, the original map and the depth image of the target house are firstly obtained. The original map is an initial map of a target house constructed according to scanning data of the laser radar. The depth image is obtained by shooting through a shooting device with a scanning distance measuring function, and the depth image comprises feature data of each room of a target house. On the basis, the obtained original map of the target house is subjected to filling correction through a preset segmentation algorithm, and a corrected house map is obtained. And analyzing and processing the depth image containing the feature data of each room of the target house to determine specific feature information and room attributes in each room. Further, after the feature information and the room attributes of each room are determined, the corrected house map is updated according to the feature information and the room attributes to obtain a complete and accurate semantic map of the target house, namely the target map, so that the accuracy of the work control of the subsequent cleaning assembly is ensured, and the cleaning effect of the house is ensured.

It can be understood that, when the cleaning assembly is building the house map through the laser radar data, because the height of the cleaning assembly is shorter, and there are a plurality of shelters in the house, lead to the unable house edge wall body data that detects of laser radar data to unable accurate reduction indoor map, make the house map of final structure incomplete, lead to the subsequent job control of cleaning assembly accurate inadequately, influence clean effect.

Therefore, according to the house map generation method provided by the invention, the acquired original map of the target house constructed by the scanning data of the laser radar is subjected to filling correction through a preset segmentation algorithm, so as to obtain a corrected regular house map. Meanwhile, the acquired depth image data containing the marker (namely feature) data of each room of the target house is analyzed and processed, and the room attribute and the feature information of each room of the target house are determined according to the processing result. And updating and displaying the corrected house map through the room attribute and the feature information, namely displaying the corresponding feature image in the corresponding room map according to the room attribute so as to obtain a semantic map (namely a target map) of the target house. Therefore, the wall data at the edge of the house can be accurately displayed in the house map, so that the indoor map can be accurately restored, a complete and accurate house semantic map is finally obtained, the accuracy of the subsequent work control of the cleaning assembly is further ensured, and the cleaning effect of the cleaning assembly on the house is improved.

In summary, in the house map generation method provided by the present invention, the obtained original map is subjected to a filling correction by a preset segmentation algorithm, so as to obtain a corrected regular house map. Meanwhile, the acquired depth image data is analyzed and processed, the room attribute and the feature information of each room of the target house are determined according to the processing result, and the corrected house map is updated according to the room attribute and the feature information to obtain the target map. Therefore, the wall data at the edge of the house can be accurately displayed in the house map, so that the indoor map can be accurately restored, a complete and accurate house semantic map is finally obtained, the accuracy of the subsequent work control of the cleaning assembly is further ensured, and the cleaning effect of the cleaning assembly on the house is improved.

The house map generation method according to the present invention may further include the following additional technical features:

in the above technical solution, correcting the original map according to a preset segmentation algorithm specifically includes: establishing a corresponding relation between an original map and a target correction result; and correcting the edge data of the original map according to the corresponding relation and a preset segmentation algorithm.

In the technical scheme, after the original map of the target house constructed by the scanning data of the laser radar is obtained, the corresponding relation between the original map and the target correction result is established based on a preset segmentation algorithm. The target correction result is an expected correction effect obtained by performing filling correction on the original map according to a preset segmentation algorithm, and the corresponding relation is an input-output relation between the original map and the target correction result. On the basis, based on the input-output relationship, the edge data of the input original map is subjected to filling correction through a preset segmentation algorithm, and the corrected house map is output. In this way, the acquired original map of the target house constructed by the scanning data of the laser radar is subjected to filling correction through a preset segmentation algorithm, so that a corrected regular house map is obtained. The wall data at the edge of the house can be accurately displayed in the house map, so that the indoor map can be accurately restored, the complete house map is finally obtained, the accuracy of the subsequent work control of the cleaning assembly is ensured, and the cleaning effect of the cleaning assembly on the house is improved.

In any of the above technical solutions, determining feature information and room attributes of a plurality of rooms of a target house according to a depth image specifically includes: carrying out target detection on the depth image according to a preset detection algorithm; and determining feature information and room attributes according to the target detection result.

In the technical scheme, after depth image data including marker (i.e. feature) data of each room of a target house is acquired, the depth image data is detected according to a preset target detection algorithm, and further specific feature information and corresponding room attributes in each room of the target house are determined according to a result of the target detection. Therefore, the feature information and the room attribute of the target house are determined based on the preset target detection algorithm, the accuracy of determining the feature information and the room attribute is guaranteed, the accuracy of subsequent map updating is further guaranteed, a complete and accurate house semantic map is obtained, the accuracy of work control of the cleaning assembly is guaranteed, and the cleaning effect is improved.

Specifically, in the actual application process, the user can set corresponding features for different rooms according to the actual decoration condition of the house. For example, features of a living room may be provided as a sofa, television, etc., features of a bedroom may be provided as a bed, features of a kitchen may be provided as a cabinet, features of a dining room may be provided as a dining table, features of a bathroom may be provided as a toilet, a wash station, etc. On the basis, the depth image data of the target house containing the feature data is detected through a preset target detection algorithm to obtain specific internal data of the target house, meanwhile, detection frame data of each feature are labeled, and then final feature information and corresponding room attributes of the target house are obtained through processing and analyzing the specific internal data and the detection frame data of the house.

In any of the above technical solutions, determining the feature information and the room attribute according to the target detection result specifically includes: training a preset excavation model according to a target detection result; and determining the room attribute and the corresponding feature information according to the trained preset mining model and the target detection result.

In the technical scheme, after a plurality of pieces of detection information are obtained by performing target detection on depth image data of a target house including feature data, a preset mining model is trained based on the plurality of pieces of detection information obtained by the detection. On the basis, the trained mining model is used for processing the plurality of detected information obtained by detection, and further room attributes and corresponding feature information of each room are output, so that each room in the target house is classified, and the corresponding feature information is determined. Therefore, the preset data mining model is trained according to the target detection result of the depth image data of the target house, and the detection result is processed through the trained mining model, so that the room attribute and the corresponding feature information of the target house are determined, the accuracy of determining the feature information and the room attribute is ensured, the accuracy of updating a subsequent map is ensured, a complete and accurate house semantic map is obtained, the accuracy of work control of the cleaning assembly is ensured, and the cleaning effect is improved.

In any of the above technical solutions, updating the corrected house map according to the feature information and the room attribute specifically includes: determining the position coordinates of the feature and the feature image according to the feature information; and updating the corrected house map according to the room attribute, the position coordinate and the feature image.

According to the technical scheme, after the room attribute and the corresponding feature information of each room of a target house are determined, the feature image of the corresponding feature and the pixel position (namely the position coordinate) of the feature in the depth image of the target house are determined according to the determined feature information, and the determined feature image is displayed at the corresponding position of the corresponding room map in the house map according to the pixel position and the corresponding room attribute. Therefore, the images of the features are correspondingly displayed at the corresponding positions in the target map based on the position coordinates and the room attributes of the features, the accuracy of displaying the images of the features is guaranteed, a complete and accurate house semantic map is generated, the accuracy of controlling the subsequent work of the cleaning assembly is guaranteed, and the cleaning effect of the cleaning assembly on houses is improved.

In any of the above technical solutions, updating the corrected house map according to the room attribute, the position coordinate, and the feature image specifically includes: determining a target room area of the corrected house map according to the room attributes; and displaying the feature images at the corresponding positions of the target room area according to the position coordinates.

In the technical scheme, after the image of the feature and the pixel position (i.e. position coordinate) of the image in the depth image of the target house are determined according to the feature information, for each room attribute, the target room area corresponding to the room name is determined from the corrected house map according to the room attribute, namely the room name. On the basis, for each feature corresponding to the room attribute, according to the pixel position (namely, the position coordinate) of the feature in the depth image of the target house, the target display position (namely, the corresponding position) of the corresponding feature in the target room area in the house map is determined, and then the image of the corresponding feature is displayed at the target display position in the house map, so that the complete and accurate house semantic map is obtained. Therefore, based on the position coordinates of the feature objects, the images of the feature objects are displayed at the corresponding positions in the corresponding room maps, the accuracy of displaying the images of the feature objects is guaranteed, a complete and accurate house semantic map is generated, the accuracy of controlling the subsequent work of the cleaning assembly is guaranteed, and the cleaning effect of the cleaning assembly on houses is improved.

In any of the above technical solutions, the obtaining of the original map of the target house specifically includes: acquiring laser radar scanning data of a target house; and generating an original map according to the scanning data of the laser radar.

In the technical scheme, when the original map of the target house is obtained, the laser radar module arranged on the cleaning assembly is used for scanning the laser radar of the target house so as to obtain the laser radar scanning data of the target house, and then the original map of the target house is constructed according to the scanning data. On the basis, the room of the constructed initial map is segmented through a segmentation algorithm, so that the original map of the target house is obtained.

According to a second aspect of the present invention, there is provided a house map generating apparatus including: the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring an original map and a depth image of a target house; the processing unit is used for correcting the original map according to a preset segmentation algorithm; the processing unit is further used for determining feature information and room attributes of a plurality of rooms of the target house according to the depth image; the processing unit is also used for updating the corrected house map according to the feature information and the room attribute to obtain a target map; wherein the feature information corresponds to the room attribute.

The invention provides a house map generation device which comprises an acquisition unit and a processing unit. When generating the house map, first, an original map of a target house and a depth image are acquired by an acquisition unit. The original map is an initial map of a target house constructed according to scanning data of the laser radar. The depth image is obtained by shooting through a shooting device with a scanning distance measuring function, and the depth image comprises feature data of each room of a target house. On the basis, the acquired original map of the target house is subjected to filling correction through a processing unit according to a preset segmentation algorithm, and a corrected house map is obtained. And analyzing and processing the depth image containing the feature data of each room of the target house to determine specific feature information and room attributes in each room. Furthermore, after the feature information and the room attributes of each room are determined, the corrected house map is updated through the processing unit according to the feature information and the room attributes to obtain a complete and accurate semantic map, namely a target map, of the target house, so that the accuracy of the work control of the subsequent cleaning assembly is ensured, and the cleaning effect of the house is ensured.

Therefore, according to the house map generation device provided by the invention, the processing unit is used for performing filling correction on the original map of the target house constructed by the laser radar scanning data acquired by the acquisition unit according to the preset segmentation algorithm so as to obtain the corrected regular house map. Meanwhile, the processing unit analyzes and processes the depth image data which is acquired by the acquisition unit and contains the marker (namely feature) data of each room of the target house, and determines the room attribute and the feature information of each room of the target house according to the processing result. And then, updating and displaying the corrected house map through the processing unit according to the room attribute and the feature information, namely displaying the corresponding feature image in the corresponding room map according to the room attribute, so as to obtain a semantic map (namely a target map) of the target house. Therefore, the wall data at the edge of the house can be accurately displayed in the house map, so that the indoor map can be accurately restored, a complete and accurate house semantic map is finally obtained, the accuracy of the subsequent work control of the cleaning assembly is further ensured, and the cleaning effect of the cleaning assembly on the house is improved.

In summary, in the house map generation device provided by the present invention, the processing unit performs the filling correction on the obtained original map through the preset segmentation algorithm, so as to obtain the corrected regular house map. Meanwhile, the processing unit analyzes and processes the depth image data acquired by the acquisition unit, determines the room attribute and the feature information of each room of the target house according to the processing result, and updates the corrected house map according to the room attribute and the feature information to obtain the target map. Therefore, the wall data at the edge of the house can be accurately displayed in the house map, so that the indoor map can be accurately restored, a complete and accurate house semantic map is finally obtained, the accuracy of the subsequent work control of the cleaning assembly is further ensured, and the cleaning effect of the cleaning assembly on the house is improved.

According to a third aspect of the present invention, there is provided a cleaning assembly comprising: a memory storing programs or instructions; and the processor is used for realizing the steps of the house map generation method in any one technical scheme when executing the program or the instructions. Therefore, the cleaning assembly provided by the third aspect of the present invention has all the advantages of the house map generation method in any one of the above-mentioned technical solutions of the first aspect, and details thereof are not repeated herein.

According to a fourth aspect of the present invention, there is provided a cleaning assembly comprising: the house map generation device in the technical scheme.

The cleaning assembly provided by the fourth aspect of the present invention includes the house map generating device in the second aspect, and therefore, the cleaning assembly has all the beneficial effects of the house map generating device in the second aspect, and details thereof are not repeated herein.

The above-described cleaning assembly according to the invention may also have the following additional technical features:

in the above technical solution, the cleaning assembly further includes: and the laser radar scanning device is used for scanning the laser radar of the target house to obtain the laser radar scanning data of the target house.

In this technical solution, the cleaning assembly further includes a laser radar scanning device. In the moving working process of the cleaning assembly, laser radar scanning is carried out on the target house through the laser radar scanning device arranged on the cleaning assembly, and therefore laser radar scanning data of the target house are obtained. On the basis, the house map generation device constructs an initial map of the target house according to the scanning data, and performs room segmentation on the constructed initial map through a segmentation algorithm to obtain an original map of the target house.

In any of the above solutions, the cleaning assembly further comprises: the system comprises a shooting device and a plurality of sensors, wherein the shooting device is used for obtaining a depth image of a target house.

In this technical scheme, above-mentioned cleaning assembly still includes camera and a plurality of sensor. Wherein, this cooperation work can realize the scanning range finding function of clean subassembly between shooting device and a plurality of sensor. In the moving process of the cleaning assembly, the target house is subjected to ranging scanning through the shooting device and the sensors arranged on the shooting device, so that the depth data of the interior of the target house, namely the depth image, is obtained.

According to a fifth aspect of the present invention, there is provided a cleaning apparatus comprising the cleaning assembly of the third aspect or the cleaning assembly of the fourth aspect. Therefore, the cleaning device proposed by the fifth aspect of the present invention has all the advantages of the cleaning assembly in the above third aspect, or the cleaning device has all the advantages of the cleaning assembly in the above fourth aspect, and details are not repeated here.

Further, the cleaning device also comprises a power supply assembly for supplying power to the cleaning assembly.

The cleaning equipment provided by the invention further comprises a power supply assembly for supplying power to the cleaning assembly in the cleaning equipment so as to ensure the normal operation of the cleaning assembly. In cleaning device's working process, the power supply unit can set up a fixed position in the house, and clean subassembly can be connected with the power supply unit, when clean subassembly execution cleaned the task, clean subassembly and power supply unit separation to clean the work according to cleaning the instruction, clean the completion back, clean the subassembly and return the power supply unit, and charge for it through the power supply unit, in order to guarantee to clean going on smoothly of task next time.

According to a sixth aspect of the present invention, there is provided a readable storage medium, on which a program or instructions are stored, the program or instructions, when executed by a processor, implementing the house map generating method according to any one of the above-mentioned aspects. Therefore, the readable storage medium provided by the sixth aspect of the present invention has all the advantages of the house map generation method in any one of the technical solutions of the first aspect, and details are not repeated herein.

Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.

Drawings

The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:

fig. 1 shows one of the flow diagrams of a house map generation method of the embodiment of the present invention;

fig. 2 is a second flowchart of a house map generation method according to an embodiment of the present invention;

fig. 3 shows a third flowchart of a house map generation method according to an embodiment of the present invention;

FIG. 4 is a fourth flowchart of a house map generation method according to an embodiment of the present invention;

fig. 5 shows a fifth flowchart of a house map generation method according to an embodiment of the present invention;

fig. 6 shows a sixth flowchart of a house map generation method of the embodiment of the present invention;

fig. 7 shows a seventh flowchart of a house map generation method of the embodiment of the present invention;

FIG. 8 shows one of the principles of a house map generation method of an embodiment of the present invention;

FIG. 9 shows a second schematic diagram of a house map generation method of an embodiment of the invention;

fig. 10 is a block diagram showing the structure of a house map generating apparatus of the embodiment of the present invention;

FIG. 11 shows one of the block diagrams of the cleaning assembly of the embodiment of the present invention;

FIG. 12 shows a second block diagram of the cleaning assembly of the embodiment of the present invention;

FIG. 13 is a block diagram showing the structure of a cleaning apparatus according to an embodiment of the present invention;

FIG. 14 shows a second block diagram of the cleaning apparatus according to the embodiment of the present invention;

fig. 15 shows a network structure diagram of the UNet model according to the embodiment of the present invention.

Detailed Description

In order that the above objects, features and advantages of the present invention can be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings. It should be noted that the embodiments of the present invention and features of the embodiments may be combined with each other without conflict.

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those specifically described herein, and therefore the scope of the present invention is not limited to the specific embodiments disclosed below.

The house map generation method and apparatus, the cleaning assembly, and the cleaning device provided in the embodiments of the present application are described in detail below with reference to fig. 1 to fig. 15 through specific embodiments and application scenarios thereof.

First embodiment, fig. 1 shows one of the flow diagrams of a house map generation method according to the embodiment of the present invention. The generation method comprises the following steps S102 to S108:

step S102, obtaining an original map and a depth image of a target house;

step S104, correcting the original map according to a preset segmentation algorithm;

step S106, determining feature information and room attributes of a plurality of rooms of a target house according to the depth image;

step S108, updating the corrected house map according to the feature information and the room attribute to obtain a target map;

wherein the feature information corresponds to the room attribute.

The execution subject of the technical scheme of the house map generation method provided by the embodiment of the invention can be a house map generation device, and can also be determined according to actual use requirements, which is not specifically limited herein. In order to more clearly describe the house map generation method provided by the present invention, the following description will be made with an execution subject of the house map generation method as a house map generation apparatus.

According to the house map generation method provided by the invention, when the house map is generated, the depth image of the target house and the original map are firstly acquired. The original map is an initial map of a target house constructed based on scanning data of the laser radar. The depth image is obtained by shooting through a shooting device with a scanning distance measuring function, and the depth image comprises feature data of each room of a target house. On the basis, the obtained original map of the target house is subjected to filling correction through a preset segmentation algorithm, and a corrected house map is obtained. And analyzing and processing the depth image containing the feature data of each room of the target house to determine specific feature information and room attributes in each room. Further, after the feature information and the room attributes of each room are determined, the corrected house map is updated according to the feature information and the room attributes to obtain a complete and accurate semantic map of the target house, namely the target map, so that the accuracy of the work control of the subsequent cleaning assembly is ensured, and the cleaning effect of the house is ensured.

When the original map is obtained, laser radar scanning can be performed on the target house through the laser radar module arranged on the cleaning assembly, so that laser radar scanning data of the target house are obtained, and initial map data are constructed according to the scanning data. On the basis, room segmentation is carried out on the constructed initial map data through a segmentation algorithm, so that the original map is obtained. In addition, in the actual application process, an initial map can be constructed through prestored laser radar scanning data to generate an original map, or the original map prestored in the storage interval of the cleaning assembly can be directly retrieved. That is to say, the house map generation method provided by the invention can generate the original map of the target house in real time through the movement track of the cleaning component for processing, and can also obtain the original map by calling the pre-stored data, and then obtain the target map by processing. For the above-mentioned obtaining manner of the original map, the user may select according to the actual situation, and is not limited in particular here.

Similarly, when obtaining above-mentioned depth image, can shoot the depth image of target house in real time through the shooting module that sets up on clean subassembly according to clean subassembly's removal orbit, can also directly call the depth image data of the target house of prestore in clean subassembly storage interval. Namely, the house map generation method provided by the invention can acquire the depth data information of the target house in real time through the moving track of the cleaning component for processing, and can also acquire the depth image by calling the pre-stored image data. For the above-mentioned obtaining manner of the depth image, the user may select the obtaining manner according to the actual situation, and is not limited specifically here.

Further, the preset segmentation algorithm may specifically be an advanced deep learning segmentation algorithm such as UNet algorithm, Solov2 algorithm, and the like. For the specific type of the preset segmentation algorithm, the user may select the type according to the actual situation, and is not limited specifically herein.

Further, the room attribute and the feature information are in a corresponding relationship with each other, and the feature information is used as identification information of a corresponding room, so that each room in the target house is identified and classified. For example, when the room attribute is a living room, the corresponding feature information is related information of a sofa and a television, when the room attribute is a bedroom, the corresponding feature information is related information of a bed, when the room attribute is a kitchen, the corresponding feature information is related information of a cabinet, when the room attribute is a dining room, the corresponding feature information is related information of a dining table, and when the room attribute is a toilet, the corresponding feature information is related information of a toilet.

In the second embodiment, fig. 2 shows a second flow chart of the house map generation method according to the second embodiment of the present invention. The generation method comprises the following steps S202 to S210:

step S202, obtaining an original map and a depth image of a target house;

step S204, establishing a corresponding relation between the original map and the target correction result;

step S206, correcting the edge data of the original map according to the corresponding relation and a preset segmentation algorithm;

step S208, determining feature information and room attributes of a plurality of rooms of the target house according to the depth image;

step S210, updating the corrected house map according to the feature information and the room attribute to obtain a target map;

wherein the feature information corresponds to the room attribute.

In this embodiment, a specific manner of correcting the original map of the target house constructed by scanning data with the laser radar is defined. Specifically, after the original map of the target house constructed by the laser radar scanning data is acquired, the corresponding relationship between the original map and the expected target correction result is established based on a preset segmentation algorithm. The target correction result is an expected correction effect obtained by performing filling correction on the original map according to a preset segmentation algorithm, and the corresponding relation is an input-output relation between the original map and the expected correction result. On the basis, based on the input-output relationship, the edge part (such as wall information) of the input original map is subjected to filling correction through a preset segmentation algorithm, and then the corrected house map is output.

In this way, the acquired original map of the target house constructed by the scanning data of the laser radar is subjected to filling correction through a preset segmentation algorithm, so that a corrected regular house map is obtained. The wall data at the edge of the house can be accurately displayed in the house map, so that the indoor map can be accurately restored, the complete house map is finally obtained, the accuracy of the subsequent work control of the cleaning assembly is ensured, and the cleaning effect of the cleaning assembly on the house is improved.

The preset segmentation algorithm may be an advanced deep learning segmentation algorithm such as UNet algorithm, Solov2 algorithm, and the like. For the specific category of the preset segmentation algorithm, the user may select the specific category according to the actual situation, and is not limited specifically herein.

For example, when the preset segmentation algorithm is the UNet algorithm (the network structure diagram of the UNet algorithm is shown in fig. 15), as shown in fig. 8, an original map 802 of a target house constructed by scanning data with a laser radar is input to a correction model constructed by the UNet algorithm, an image of an edge portion of the input original map 802 is subjected to a complementary correction by the correction model, and a house map 804 of a corrected rule is output.

In a third embodiment, fig. 3 shows a third flow chart of the house map generation method according to the third embodiment of the present invention. In this embodiment, the cleaning assembly is provided with an inlet valve and an outlet valve, and the generating method comprises the following steps S302 to S310:

step S302, acquiring an original map and a depth image of a target house;

step S304, correcting the original map according to a preset segmentation algorithm;

step S306, carrying out target detection on the depth image according to a preset detection algorithm;

step S308, determining feature information and room attributes of a plurality of rooms of a target house according to a target detection result;

step S310, updating the corrected house map according to the feature information and the room attribute to obtain a target map;

wherein the feature information corresponds to the room attribute.

In this embodiment, a specific manner of determining the room attribute and the feature information of each room of the target house by analyzing the depth image data including the feature data of each room of the target house is defined. Specifically, after obtaining depth image data including marker (i.e., feature) data of each room of a target house, the depth image data is detected according to a preset target detection algorithm, and further, according to a result of the target detection, specific feature information and corresponding room attributes in each room of the target house are determined. Therefore, the feature information and the room attribute of the target house are determined based on the preset target detection algorithm, the accuracy of determining the feature information and the room attribute is guaranteed, the accuracy of subsequent map updating is further guaranteed, a complete and accurate house semantic map is obtained, the accuracy of work control of the cleaning assembly is guaranteed, and the cleaning effect is improved.

It should be noted that, the user may select and set the feature objects of each room according to the actual decoration condition of the house, and may also select and set the feature objects of each room according to the priori knowledge. For the specific manner of selecting and setting the feature, the user may select the feature according to the actual situation, and is not limited specifically here.

Further, the preset detection algorithm may specifically be a target detection algorithm such as a centeret algorithm, a Yolov5 algorithm, and the like. For the specific type of the preset detection algorithm, the user may select the detection algorithm according to the actual situation, and is not limited specifically herein.

Further, the target detection result may specifically include detection information such as x, y, w, h, label, socre, slam _ x, slam _ y, square, num, and target. Wherein x is an abscissa of the feature detection frame, y is an ordinate of the feature detection frame, w is a width of the feature detection frame, h is a height of the feature detection frame, label is a category of the feature detection frame, slam _ x is an abscissa of a pose of the cleaning component in the house map, slam _ y is an ordinate of a pose of the cleaning component in the house map, square is a room area, num is a number of rooms, and target is a room attribute (or a room name). On the basis, all or part of the detection information is processed and analyzed to identify the room attribute and the corresponding feature information of each room.

In a fourth embodiment, fig. 4 shows a fourth flowchart of the house map generation method according to the embodiment of the present invention. The generation method comprises the following steps S402 to S412:

step S402, obtaining an original map and a depth image of a target house;

step S404, correcting an original map according to a preset segmentation algorithm;

step S406, carrying out target detection on the depth image according to a preset detection algorithm;

step S408, training a preset excavation model according to a target detection result;

step S410, according to the trained preset mining model and a target detection result, determining room attributes of a plurality of rooms of a target house and corresponding feature information;

and step S412, updating the corrected house map according to the feature information and the room attribute to obtain a target map.

In this embodiment, a specific manner of identifying the room attribute and the corresponding feature information of each room of the target house by the processing and analysis of the target detection result of the depth image data is defined. Specifically, after a plurality of pieces of detection information are obtained by performing target detection on depth image data of a target house including feature data, a preset mining model is trained based on the plurality of pieces of detection information obtained by the detection. On the basis, the trained mining model is used for processing the plurality of detected information obtained by detection, and further room attributes and corresponding feature information of each room are output, so that each room in the target house is classified, and the corresponding feature information is determined. Therefore, the preset data mining model is trained and optimized according to the target detection result of the depth image data of the target house, and then the detection result is processed through the trained mining model, so that the room attribute and the corresponding feature information of the target house are determined, the accuracy of determining the feature information and the room attribute is ensured, the accuracy of updating a subsequent map is ensured, a complete and accurate house semantic map is obtained, the accuracy of work control of the cleaning assembly is ensured, and the cleaning effect is improved.

The target detection result may include detection information such as an abscissa x of the feature detection box, an ordinate y of the feature detection box, a width w of the feature detection box, a height h of the feature detection box, a category label of the feature detection box, an abscissa slam _ x of a pose of the cleaning component in the house map, an ordinate slam _ y of a pose of the cleaning component in the house map, a room area square, a room number num, a room attribute (or a room name) target, and a total number xn of different types of detection boxes in the room. On the basis, a preset data mining model is trained through all or part of the detection information in the detection information, and the detection information is processed through the trained mining model to determine the room attribute and the corresponding feature information of each room.

Further, the preset mining model may be a data mining model such as an Xgboost model, a Catboost model, and a Lightgbm model. For the specific type of the preset mining model, the user may select the type according to the actual situation, and is not limited specifically here.

Illustratively, in the case that the preset mining model is an Xgboost model and the number of rooms is 5, the Xgboost model may be trained by selecting the detection information x0, x1, x2, x3, x4, x5, square, num and target. Wherein x0-x5 are the total number of different types of detection boxes in each room, square is the area of the room, num is the number of the rooms, and target is the attribute (or name) of the room.

On the basis, the Xgboost model can be trained by the following training parameters and corresponding parameter values in the following Table 1:

table 1: training parameter table

It should be noted that the training parameters and the corresponding parameter values in the training parameter table are only examples in a specific embodiment, and in an actual application process, the training parameters and the corresponding parameter values may be set according to an actual situation, which is not limited herein.

Fifth embodiment, fig. 5 shows a fifth flowchart of the house map generation method according to the fifth embodiment of the present invention. In this embodiment, the working temperature includes an ambient temperature of the cleaning assembly, and the generating method includes the following steps S502 and S510:

step S502, obtaining an original map and a depth image of a target house;

step S504, correcting the original map according to a preset segmentation algorithm;

step S506, determining feature information and room attributes of a plurality of rooms of a target house according to the depth image;

step S508, determining the position coordinates of the feature and the feature image according to the feature information;

step S510, updating the corrected house map according to the room attribute, the position coordinate and the feature image to obtain a target map;

wherein the feature information corresponds to the room attribute.

In this embodiment, a specific manner of updating the corrected house map based on the feature information and the room attribute is defined. Specifically, after the room attribute of each room of the target house and the corresponding feature information are determined, the feature image of the corresponding feature and the pixel position (i.e., the position coordinate) of the feature in the depth image of the target house are determined according to the determined feature information, and then the determined feature image is displayed at the corresponding position of the corresponding room map in the house map according to the pixel position and the corresponding room attribute. Therefore, the images of the features are correspondingly displayed at the corresponding positions in the target map based on the position coordinates and the room attributes of the features, the accuracy of displaying the images of the features is guaranteed, a complete and accurate house semantic map is generated, the accuracy of controlling the subsequent work of the cleaning assembly is guaranteed, and the cleaning effect of the cleaning assembly on houses is improved.

Sixth embodiment, fig. 6 shows a sixth flowchart of a house map generation method according to the sixth embodiment of the present invention. The generation method comprises the following steps S602 to S612:

step S602, obtaining an original map and a depth image of a target house;

step S604, correcting the original map according to a preset segmentation algorithm;

step S606, determining feature information and room attributes of a plurality of rooms of a target house according to the depth image;

step S608, determining the position coordinates of the feature and the feature image according to the feature information;

step S610, determining a target room area of the corrected house map according to the room attribute;

step S612, displaying the feature object image at the corresponding position of the target room area according to the position coordinates to obtain a target map;

wherein the feature information corresponds to the room attribute.

In this embodiment, a specific manner of updating and displaying the corrected house map based on the pixel position and the room attribute of the feature based on the image of the feature is defined. Specifically, after the image of the feature and the pixel position (i.e., the position coordinates) thereof in the depth image of the target house are determined based on the feature information, for each room attribute, a target room area corresponding to the room name is determined from the corrected house map based on the room attribute, i.e., the room name. On the basis, for each feature corresponding to the room attribute, according to the pixel position (namely, the position coordinate) of the feature in the depth image of the target house, the target display position (namely, the corresponding position) of the corresponding feature in the target room area in the house map is determined, and then the image of the corresponding feature is displayed at the target display position in the house map, so that the complete and accurate house semantic map is obtained. Therefore, based on the position coordinates of the feature objects, the images of the feature objects are displayed at the corresponding positions in the corresponding room maps, the accuracy of displaying the images of the feature objects is guaranteed, a complete and accurate house semantic map is generated, the accuracy of controlling the subsequent work of the cleaning assembly is guaranteed, and the cleaning effect of the cleaning assembly on houses is improved.

Exemplarily, fig. 9 shows a display map of a feature image of an embodiment of the present invention. As shown in fig. 9 (a), the corrected house map is divided into seven rooms in total of three bedrooms, one toilet, one dining room, one living room, and one kitchen. On the basis, the trained data mining model is used for analyzing and processing the target detection result of the depth map data of each room so as to identify the room attribute of each room and the feature information in the room attribute. On this basis, as shown in (b) in fig. 9, the images of the respective features are displayed at the respective positions in the corresponding room map in accordance with the position coordinates of the features. Wherein, the characteristic objects of each room are arranged as follows: the characteristic objects of the bedroom are a bed, the characteristic objects of the bathroom are a closestool and a wash platform, the characteristic objects of the dining room are a dining table, the characteristic objects of the living room are a television and a sofa, and the characteristic objects of the kitchen are a cabinet.

Seventh embodiment, fig. 7 shows a seventh flowchart of a house map generation method according to an embodiment of the present invention. In this embodiment, the working temperature includes a tank temperature of the cleaning assembly, and the generating method includes the following steps S702 to S710:

step S702, acquiring laser radar scanning data of a target house;

step S704, generating an original map according to the laser radar scanning data;

step S706, correcting the original map according to a preset segmentation algorithm;

step 708, determining feature information and room attributes of a plurality of rooms of the target house according to the depth image of the target house;

step S710, updating the corrected house map according to the feature information and the room attribute to obtain a target map;

wherein the feature information corresponds to the room attribute.

In this embodiment, the specific manner of acquiring the original map is defined. Specifically, during the original map, the laser radar module arranged on the cleaning component is used for scanning the laser radar of the target house to obtain the laser radar scanning data of the target house, and then the initial image data is constructed according to the scanning data. On the basis, room segmentation is carried out on the constructed initial image data through a segmentation algorithm, so that an original map of the target house is obtained.

In addition, in the actual application process, an initial map can be constructed through prestored laser radar scanning data to generate an original map, or the original map prestored in the storage interval of the cleaning assembly can be directly retrieved. That is to say, the house map generation method provided by the invention can generate the original map of the target house in real time through the movement track of the cleaning component for processing, and can also obtain the original map by calling the pre-stored data, and then obtain the target map by processing. For the above-mentioned obtaining manner of the original map, the user may select according to the actual situation, and is not limited in particular here.

Eighth embodiment, fig. 10 is a block diagram showing a structure of a house map generation apparatus 1000 according to an embodiment of the present invention. The house map generation apparatus 1000 includes an acquisition unit 1002 and a processing unit 1004:

an obtaining unit 1002, configured to obtain an original map and a depth image of a target house;

a processing unit 1004 for correcting the original map according to a preset segmentation algorithm;

the processing unit 1004 is further used for determining feature information and room attributes of a plurality of rooms of the target house according to the depth image;

the processing unit 1004 is further configured to update the corrected house map according to the feature information and the room attribute, so as to obtain a target map;

wherein the feature information corresponds to the room attribute.

The house map generation apparatus 1000 according to the embodiment of the present invention includes an acquisition unit 1002 and a processing unit 1004. In generating the house map, first, a depth image of a target house and an original map are acquired by the acquisition unit 1002. The original map is an initial map of a target house constructed according to scanning data of the laser radar. The depth image is obtained by shooting through a shooting device with a scanning distance measuring function, and the depth image comprises feature data of each room of a target house. On the basis, the processing unit 1004 performs the filling correction on the obtained original map of the target house according to a preset segmentation algorithm to obtain a corrected house map. And analyzing and processing the depth image containing the feature data of each room of the target house to determine specific feature information and room attributes in each room. Further, after the feature information and the room attribute of each room are determined, the corrected house map is updated by the processing unit 1004 according to the feature information and the room attribute, so as to obtain a complete and accurate semantic map, i.e. a target map, of the target house, so as to ensure the accuracy of the work control of the subsequent cleaning component, thereby ensuring the cleaning effect of the house.

Therefore, the house map generation apparatus 1000 according to the present invention performs, by the processing unit 1004, a supplementary correction on the original map of the target house constructed by the lidar scanning data acquired by the acquisition unit 1002 according to a preset segmentation algorithm, so as to obtain a corrected regular house map. Meanwhile, the depth image data including the marker (i.e., feature) data of each room of the target house acquired by the acquisition unit 1002 is analyzed and processed by the processing unit 1004, and the room attribute and the feature information of each room of the target house are determined according to the processing result. The processing unit 1004 further updates and displays the modified house map according to the room attributes and the feature information, that is, displays the corresponding feature image in the corresponding room map according to the room attributes, thereby obtaining the semantic map (i.e., the target map) of the target house. Therefore, the wall data at the edge of the house can be accurately displayed in the house map, so that the indoor map can be accurately restored, a complete and accurate house semantic map is finally obtained, the accuracy of the subsequent work control of the cleaning assembly is further ensured, and the cleaning effect of the cleaning assembly on the house is improved.

Wherein, when acquireing above-mentioned original map, can carry out the laser radar scanning to the target house through the laser radar module that sets up on clean subassembly, the laser radar scanning data of acquisition unit 1002 acquisition target house. On the basis of the original image data, the processing unit 1004 performs room segmentation on the constructed original image data by a segmentation algorithm according to the initial image data constructed by the scanning data, so as to obtain the original map. In addition, in an actual application process, the laser radar scanning data prestored in the storage interval of the cleaning assembly can be acquired through the acquisition unit 1002, and then the processing unit 1004 constructs an initial map to generate an original map, or the acquisition unit 1002 directly retrieves the original map prestored in the storage interval of the cleaning assembly. That is to say, the house map generation method provided by the invention can generate the original map of the target house in real time through the movement track of the cleaning component for processing, and can also obtain the original map by calling the pre-stored data, and then obtain the target map by processing. For the above-mentioned obtaining manner of the original map, the user may select according to the actual situation, and is not limited in particular here.

Similarly, when the depth image is obtained, the depth image of the target house can be shot in real time through the shooting module arranged on the cleaning component according to the moving track of the cleaning component, and the depth image data of the target house prestored in the storage interval of the cleaning component can also be directly called through the obtaining unit 1002. Namely, the house map generation method provided by the invention can acquire the depth data of the target house in real time through the moving track of the cleaning component for processing, and can also acquire the depth image by calling the pre-stored image data. For the above-mentioned obtaining manner of the depth image, the user may select the obtaining manner according to the actual situation, and is not limited specifically here.

Furthermore, the feature information and the room attributes are in a corresponding relationship with each other, and the feature information is used as identification information of a corresponding room to identify and classify each room in the target house. For example, when the room attribute is a living room, the corresponding feature information is related information of a sofa and a television, when the room attribute is a bedroom, the corresponding feature information is related information of a bed, when the room attribute is a kitchen, the corresponding feature information is related information of a cabinet, when the room attribute is a dining room, the corresponding feature information is related information of a dining table, and when the room attribute is a toilet, the corresponding feature information is related information of a toilet.

In summary, in the house map generation apparatus 1000 according to the present invention, the processing unit 1004 performs a supplementary correction on the acquired original map by using a preset segmentation algorithm, so as to obtain a corrected regular house map. Meanwhile, the processing unit 1004 analyzes the depth image data acquired by the acquisition unit 1002, determines the room attribute and the feature information of each room of the target house according to the processing result, and updates the corrected house map according to the room attribute and the feature information to obtain the target map. Therefore, the wall data at the edge of the house can be accurately displayed in the house map, so that the indoor map can be accurately restored, a complete and accurate house semantic map is finally obtained, the accuracy of the subsequent work control of the cleaning assembly is further ensured, and the cleaning effect of the cleaning assembly on the house is improved.

In this embodiment, further, the processing unit 1004 is specifically configured to: establishing a corresponding relation between an original map and a target correction result; and correcting the edge data of the original map according to the corresponding relation and a preset segmentation algorithm.

In this embodiment, further, the processing unit 1004 is specifically configured to: carrying out target detection on the depth image according to a preset detection algorithm; and determining feature information and room attributes according to the target detection result.

In this embodiment, further, the processing unit 1004 is specifically configured to: training a preset excavation model according to a target detection result; and determining the room attribute and the corresponding feature information according to the trained preset mining model and the target detection result.

In this embodiment, further, the processing unit 1004 is specifically configured to: determining the position coordinates of the feature and the feature image according to the feature information; and updating the corrected house map according to the room attribute, the position coordinate and the feature image.

In this embodiment, further, the processing unit 1004 is specifically configured to: determining a target room area of the corrected house map according to the room attributes; and displaying the feature images at the corresponding positions of the target room area according to the position coordinates.

In this embodiment, further, the obtaining unit 1002 is specifically configured to: acquiring laser radar scanning data of a target house; the processing unit 1004 may be specifically configured to: and generating an original map according to the scanning data of the laser radar.

In the ninth embodiment, fig. 11 is a block diagram illustrating a cleaning assembly 1100 provided in an embodiment of the present invention. Wherein, this cleaning assembly 1100 includes:

a memory 1102, the memory 1102 having stored thereon programs or instructions;

a processor 1104, the processor 1104 implementing the steps of the house map generation method in any of the above embodiments when the processor 1104 executes the above programs or instructions.

The cleaning assembly 1100 provided in this embodiment includes the memory 1102 and the processor 1104, and when the program or the instructions in the memory 1102 are executed by the processor 1104, the steps of the house map generation method in any of the above embodiments are implemented, so that the cleaning assembly 1100 has all the beneficial effects of the house map generation method in any of the above embodiments, and details are not described here again.

In particular, the memory 1102 and the processor 1104 may be connected by a bus or other means. The Processor 1104 may include one or more Processing units, and the Processor 1104 may be a Central Processing Unit (CPU), a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA), or the like.

Tenth embodiment, fig. 12 shows a schematic structural diagram of a cleaning assembly 1200 according to an embodiment of the present invention, where the cleaning assembly 1200 includes the house map generating device 1000 in the foregoing embodiment, and therefore, the cleaning assembly 1200 has all the beneficial effects of the house map generating device 1000 in any of the foregoing embodiments, which is not described herein again.

In this embodiment, the cleaning assembly 1200 further comprises a laser radar scanning device. During the moving operation of the cleaning assembly 1200, the laser radar scanning device disposed thereon scans the target house to obtain the laser radar scanning data of the target house. On this basis, the house map generating apparatus 1000 constructs initial image data from the scan data, and performs room segmentation on the constructed initial image data by a segmentation algorithm to obtain an original map of the target house.

In this embodiment, the cleaning assembly 1200 further includes a plurality of sensors and a camera. Wherein, this shooting device and the cooperation work between a plurality of sensors can realize cleaning assembly 1200's scanning range finding function. During the moving operation of the cleaning assembly 1200, the target house is scanned by the camera and the sensors arranged thereon, so as to obtain the depth data of the inside of the target house, i.e. the depth image.

Specifically, the camera may be an RGB camera, an RGB-D (depth) camera, or the like, and the plurality of sensors may include an infrared sensor, a temperature sensor, a distance sensor, or the like. The specific types of the camera and the plurality of sensors in the cleaning assembly 1200 can be selected according to actual situations, and are not particularly limited.

Eleventh, fig. 13 is a block diagram illustrating a structure of a cleaning apparatus 1300 according to an embodiment of the present invention. Wherein the cleaning apparatus 1300 comprises: the cleaning assembly 1100 of the above embodiment. Therefore, the cleaning apparatus 1300 has all the technical effects of the cleaning assembly 1100 in the above embodiments, and will not be described herein.

In the twelfth embodiment, fig. 14 is a block diagram illustrating a structure of a cleaning apparatus 1400 according to an embodiment of the present invention. Wherein the cleaning apparatus 1400 comprises: the cleaning assembly 1200 of the above embodiment. Therefore, the cleaning apparatus 1400 has all the technical effects of the cleaning assembly 1200 in the above embodiments, and will not be described herein.

According to the cleaning device provided by the embodiment of the invention, further, the cleaning device further comprises a power supply assembly for supplying power to the cleaning assembly in the cleaning device so as to ensure the normal operation of the cleaning assembly. In cleaning device's working process, the power supply unit can set up a fixed position in the house, and clean subassembly can be connected with the power supply unit, when clean subassembly execution cleaned the task, clean subassembly and power supply unit separation to clean the work according to cleaning the instruction, clean the completion back, clean the subassembly and return the power supply unit, and charge for it through the power supply unit, in order to guarantee to clean going on smoothly of task next time.

Thirteenth, an embodiment of the sixth aspect of the present invention, provides a readable storage medium. On which a program or instructions are stored which, when executed by a processor, implement the steps of the house map generation method as in any of the embodiments described above.

The readable storage medium provided by the embodiment of the present invention stores a program or instructions, and when the program or instructions are executed by a processor, the steps of the house map generation method in any of the above embodiments can be implemented. Therefore, the readable storage medium has all the advantages of the house map generation method in any of the embodiments described above, and details are not described herein. In particular, the readable storage medium may include any medium capable of storing or transmitting information. Examples of readable storage media include electronic circuits, semiconductor Memory devices, Read-Only memories (ROMs), Random Access Memories (RAMs), Compact Disc Read-Only memories (CD-ROMs), flash memories, erasable ROMs (eroms), magnetic tapes, floppy disks, optical disks, hard disks, optical fiber media, Radio Frequency (RF) links, optical data storage devices, and so forth. The code segments may be downloaded via computer networks such as the internet, intranets, etc.

In the description herein, the terms "first" and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance unless explicitly stated or limited otherwise; the terms "connected," "mounted," "secured," and the like are to be construed broadly and include, for example, fixed connections, removable connections, or integral connections; may be directly connected or indirectly connected through an intermediate. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.

In the description herein, the description of the terms "one embodiment," "some embodiments," "specific embodiments," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

In addition, the technical solutions in the embodiments of the present invention may be combined with each other, but it must be based on the realization of those skilled in the art, and when the technical solutions are contradictory or cannot be realized, such a combination of technical solutions should not be considered to exist, and is not within the protection scope of the present invention.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A house map generation method is characterized by comprising the following steps:

acquiring an original map and a depth image of a target house;

correcting the original map according to a preset segmentation algorithm;

determining feature information and room attributes of a plurality of rooms of the target house according to the depth image;

updating the corrected house map according to the feature information and the room attribute to obtain a target map;

wherein the feature information corresponds to the room attribute.

2. The house map generation method according to claim 1, wherein the correcting the original map according to a preset segmentation algorithm specifically includes:

establishing a corresponding relation between the original map and a target correction result;

and correcting the edge data of the original map according to the corresponding relation and a preset segmentation algorithm.

3. The house map generation method according to claim 1, wherein the determining feature information and room attributes of the plurality of rooms of the target house according to the depth image specifically includes:

carrying out target detection on the depth image according to a preset detection algorithm;

and determining the feature information and the room attribute according to the target detection result.

4. The house map generation method according to claim 3, wherein the determining the feature information and the room attribute according to the target detection result specifically includes:

training a preset excavation model according to the target detection result;

and determining the room attribute and the corresponding feature information according to the trained preset mining model and the target detection result.

5. The house map generation method according to claim 1, wherein the updating of the corrected house map according to the feature information and the room attribute specifically includes:

determining the position coordinates of the feature and the feature image according to the feature information;

and updating the corrected house map according to the room attribute, the position coordinate and the feature image.

6. The house map generation method according to claim 5, wherein the updating of the corrected house map based on the room attribute, the position coordinate, and the feature image specifically includes:

determining a target room area of the corrected house map according to the room attribute;

and displaying the feature images at corresponding positions of the target room area according to the position coordinates.

7. The house map generation method according to any one of claims 1 to 6, wherein the obtaining of the original map of the target house specifically includes:

acquiring laser radar scanning data of the target house;

and generating the original map according to the laser radar scanning data.

8. A house map generation apparatus, characterized by comprising:

the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring an original map and a depth image of a target house;

the processing unit is used for correcting the original map according to a preset segmentation algorithm;

the processing unit is further used for determining feature information and room attributes of a plurality of rooms of the target house according to the depth image;

the processing unit is further used for updating the corrected house map according to the feature information and the room attribute to obtain a target map;

wherein the feature information corresponds to the room attribute.

9. A cleaning assembly, comprising:

a memory storing programs or instructions;

a processor which, when executing the program or instructions, carries out the steps of the premises map generation method of any of claims 1 to 7.

10. A cleaning assembly, comprising:

the house map generating apparatus of claim 8.

11. The cleaning assembly of claim 10, further comprising:

and the laser radar scanning device is used for scanning the laser radar of the target house to obtain the laser radar scanning data of the target house.

12. The cleaning assembly of claim 10, further comprising:

the system comprises a shooting device and a plurality of sensors, wherein the shooting device is used for obtaining a depth image of a target house.

13. A cleaning apparatus, comprising:

a cleaning assembly as claimed in any one of claims 9 to 12.

14. The cleaning apparatus defined in claim 13, further comprising:

and the power supply assembly is used for supplying power to the cleaning assembly.

15. A readable storage medium on which a program or instructions are stored, characterized in that the program or instructions, when executed by a processor, implement the steps of the house map generation method according to any one of claims 1 to 7.