Dust display system and method based on polychromatic light and self-cleaning equipment

Abstract

The embodiment of the invention provides a dust display system and method based on polychromatic light and autonomous cleaning equipment, and belongs to the field of intelligent cleaning equipment. The system comprises a multi-color detection light emitter which is configured to project at least two detection light beams with different wavelengths to a to-be-cleaned ground area in front of or laterally in front of the autonomous cleaning equipment; the light receiver is configured to receive a light signal from the ground area to be cleaned and generate corresponding image data; the control module is electrically connected with the multi-color detection light emitter and the light receiver, and is configured to control the multi-color detection light emitter to switch and emit detection light beams with different wavelengths in a time-sharing mode; and on the basis of multiple groups of image data acquired by the light receiver under the irradiation of the detection light beams with different wavelengths, smudginess identification is performed, and a cleaning strategy control signal is output. The ground background and real dirt can be effectively distinguished, and the false alarm rate is greatly reduced; and meanwhile, the identification capability of smudginess colors and properties is improved, and the method is suitable for diversified family smudginess scenes.

Description

Technical Field

[0001] This invention relates to the field of intelligent cleaning equipment technology, specifically to a dust detection system, method, and autonomous cleaning equipment based on multicolor light. Background Technology

[0002] With technological advancements, modern cleaning appliances such as robotic vacuums, mops, and regular vacuums are evolving towards greater intelligence and automation. Among these advancements, the ability to detect dirt and grime on the floor and perform targeted cleaning is key to improving cleaning efficiency and effectiveness. Currently, some high-end products on the market are equipped with a "dust detection" function, typically using green laser or LED light sources to project parallel beams onto the ground. Because the human eye is most sensitive to green light, dust and solid particles on the ground will form distinct shadows under green light, making them visible to the user.

[0003] However, the aforementioned existing technologies have many limitations: First, it was originally designed for "human vision" rather than machine vision recognition. The image sensors (such as CMOS) of robotic vacuum cleaners have different response characteristics to different wavelengths of light than the human eye. For example, most CMOS sensors are more sensitive to red light than the human eye, which means that a single green light solution cannot fully utilize the sensor's performance and results in poor detection.

[0004] Secondly, monochromatic light detection is highly susceptible to interference from complex ground backgrounds. When the ground itself has patterns, color blocks, or color variations, monochromatic light struggles to effectively distinguish the background from actual dirt, leading to misidentification. Furthermore, for dirt that is similar in color to the background (such as pet feces on a brown carpet) or liquid dirt with some translucency (such as pale yellow urine), the single-dimensional optical information provided by monochromatic light (such as shadows or reflection intensity) is insufficient for accurate identification, easily resulting in missed detections or misjudgments.

[0005] Therefore, there is an urgent need for a detection solution that can effectively adapt to complex home floor environments and accurately distinguish between solid and liquid dirt of different colors, in order to improve the intelligence level and cleaning effect of cleaning equipment. Summary of the Invention

[0006] The purpose of this invention is to provide a multi-color light-based dust detection system, method, and self-cleaning device that can effectively distinguish between the ground background and actual dirt, greatly reducing the false alarm rate; at the same time, it improves the ability to identify the color and nature of dirt, and is suitable for diverse household dirt scenarios.

[0007] To achieve the above objectives, embodiments of the present invention provide a multicolor light-based dust detection system, the dust detection system comprising: A multi-color detection light emitter is configured to project at least two different wavelengths of detection light beams onto the area of ​​the floor to be cleaned in front of or to the side of the autonomous cleaning device. A light receiver, configured to receive light signals from the area of ​​the floor to be cleaned and generate corresponding image data; A control module, electrically connected to the multicolor detection light emitter and the light receiver, is configured to: The multicolor detector light emitter is controlled to switch between emitting different wavelengths of the detector light beam in a time-division manner; Based on multiple sets of image data acquired by the optical receiver under different wavelength detection beam illumination, dirt is identified, and a cleaning strategy control signal is output to the autonomous cleaning device.

[0008] Preferably, the multicolor detection light emitter includes multiple independently controllable light-emitting units, which are selected from light-emitting diodes (LEDs) or laser diodes (LDs), and the wavelength combination emitted includes at least two of red light, green light, and blue light.

[0009] Preferably, the multicolor detection light emitter further includes optical elements configured to shape the detection beam into parallel or linear light to cover the area of ​​the ground to be cleaned.

[0010] Preferably, the light receiver is a CMOS image sensor.

[0011] Preferably, the control module performs dirt identification by: For solid stains, by comparing the shape, position and contour consistency of shadows in images under different wavelength detection beams, and combining the spectral characteristics of the reflected light from solid stains on the dirty surface, the background pattern on the ground is distinguished from the actual dirt, and the color of the solid stain is determined. For liquid stains, the nature and color of the liquid stain are determined by analyzing the intensity distribution and color information of the reflected light from the liquid stain in images under different wavelength detection beams.

[0012] Preferably, the wavelength emitted by the multicolor detection light emitter also includes an invisible light band, which includes ultraviolet light for sterilizing and / or infrared light for drying dirty areas.

[0013] Preferably, the control module is further configured to: after identifying specific biological contamination, control the multicolor detection light emitter to emit ultraviolet light for irradiation and sterilization, or to emit infrared light for drying.

[0014] On the other hand, the present invention provides an autonomous cleaning device, including a device body, a moving mechanism, a cleaning mechanism and a control system, wherein the control system integrates the aforementioned dust detection system based on multicolor light.

[0015] In another aspect, the present invention provides a dust detection method based on multicolor light, which uses the above-described system for dust detection, including: The multicolor detection light emitter is controlled to illuminate the ground area to be detected with a first wavelength, and the first image data is acquired through the light receiver. The multicolor detection light emitter is controlled to switch to the second wavelength to illuminate the ground area to be detected, and the second image data is acquired through the light receiver. By fusing and analyzing the first image data and the second image data, the type, color, and distribution of dirt on the ground can be identified; Based on the identification results, generate and execute corresponding cleaning strategies.

[0016] Preferably, the fusion analysis includes: Calculate the difference images between images of different wavelengths to eliminate static background interference; Extract the feature regions in the difference image and analyze the gray values ​​or RGB component ratios of the feature regions under different wavelength images to determine the color attributes of the dirt. Based on the texture, edge features, and reflective properties of the aforementioned feature regions, solid stains and liquid stains can be distinguished.

[0017] The above technical solution involves firstly, the control module driving a multi-color detector light emitter to cycle through or switch between different colors of light at a high frequency. For the same ground area, the light receiver rapidly acquires multiple frames of images under different colored light, such as red, green, and blue. The control module then fuses and analyzes this multispectral image sequence. For solid stain identification: Solid particles block light and create shadows. While the color of the background pattern may change under different colors of light, the position and outline of the shadow of a real stain remain stable. Image differencing and other techniques can highlight these stable shadow features, effectively filtering out background interference. Furthermore, the reflectivity of stain surfaces varies with different colors of light; analyzing their reflectance spectrum can further determine the color of the stain (e.g., distinguishing between dark soil and light-colored paper scraps).

[0018] For liquid stain identification: Liquids typically exhibit specular reflection. Under different colored light, the intensity and color of the reflected light in a liquid area directly reveal the liquid's properties. For example, clear water strongly reflects all colors of light, while pale yellow urine reflects blue light less strongly. By analyzing the "color" of the reflected light, the type of liquid can be identified.

[0019] Based on the identification results, the control module can instruct the cleaning equipment to execute better cleaning strategies, such as repeatedly cleaning heavily soiled areas, reducing the travel speed, activating the mopping module to increase the water output, and even activating the ultraviolet sterilization function for pet urine areas and using infrared drying after cleaning.

[0020] This system utilizes multi-color light information to effectively distinguish between the ground background and actual dirt, greatly reducing the false alarm rate and achieving high accuracy. Furthermore, its ability to identify the color and nature of dirt allows it to handle a wider variety of household dirt scenarios, demonstrating high robustness.

[0021] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0022] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a dust detection system based on multicolor light according to one embodiment of the present invention; Figure 2 This is a schematic flowchart of a dust detection method based on multicolor light according to one embodiment of the present invention.

[0023] Explanation of reference numerals in the attached figures 1-Self-cleaning equipment, 2-Multi-color detection light emitter, 3-Light receiver, 4-Detection beam, 5-Liquid stain, 6-Reflected light from liquid stain, 7-Solid stain, 8-Reflected light from solid stain, 9-Shadow. Detailed Implementation

[0024] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0025] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application all comply with relevant laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0026] See Figure 1 The present invention provides a dust detection system based on multicolor light, the dust detection system comprising: Multicolor detection light emitter 2 is configured to project at least two different wavelength detection light beams 4 onto the ground area to be cleaned in front of or to the side of the autonomous cleaning device 1. Light receiver 3 is configured to receive light signals from the area of ​​the floor to be cleaned and generate corresponding image data; The control module is electrically connected to the multicolor detection light emitter 2 and the light receiver 3, and is configured as follows: Control the multicolor detection light emitter 2 to switch the emission of detection light beams 4 of different wavelengths in a time-division manner; Based on multiple sets of image data acquired by the optical receiver 3 under the illumination of detection beams of different wavelengths, dirt is identified, and cleaning strategy control signals are output to the autonomous cleaning device 1.

[0027] In this multicolor light-based dust detection system, the multicolor detection light emitter 2 serves as the illumination source, its core function being the ability to generate light of two or more different wavelengths. In a preferred embodiment, the multicolor detection light emitter 2 comprises multiple independently controllable light-emitting units, which can be light-emitting diodes (LEDs) or laser diodes (LDs). However, considering cost, maturity, and sensor compatibility, the wavelength combination preferably includes at least two of red light (center wavelength 620-650nm), green light (center wavelength 520-550nm), and blue light (center wavelength 450-480nm). For example, an RGB tricolor LED array can be used.

[0028] To create an effective illumination area, the multicolor detection light emitter 2 also integrates optical elements such as convex lenses, cylindrical mirrors, or gratings. These optical elements shape the light emitted by the light-emitting unit, for example, by diffusing or collimating a point light source into a parallel beam, or focusing it into a clear linear light band. The purpose of this is to ensure that the detection beam can cover a sufficiently wide and uniformly illuminated area in front of the cleaning equipment, ensuring effective scanning of dirt along the path.

[0029] In this embodiment, the light receiver 3 is essentially an image sensor, installed adjacent to the multicolor detection light emitter 2 and facing the same detection area. Preferably, the light receiver 3 employs a complementary metal-oxide-semiconductor (CMOS) image sensor, which offers advantages such as low cost, high integration, low power consumption, and fast response speed, making it ideal for consumer electronics. Alternatively, a charge-coupled device (CCD) sensor can also be used as an alternative.

[0030] The viewing angle and resolution of the light receiver 3 need to match the coverage area of ​​the probe beam 4 to ensure that the details of the ground can be clearly captured. When the probe beam 4 shines on the ground, the light will interact with the ground background, solid dirt and liquid dirt in different ways (such as reflection, scattering, and occlusion). The light receiver 3 captures these light signals carrying scene information and converts them into digital image data for subsequent processing.

[0031] In this system, the control module is the core responsible for coordinating and controlling the entire detection and identification process. It can be the main control microprocessor (MCU) of the cleaning equipment or a dedicated image processing chip (such as a DSP, FPGA, or ASIC). It is programmed to perform the following functions: First, it controls the multi-color detection light emitter 2 to switch the emission of detection beams of different wavelengths in a time-division multiplexing manner. That is, within extremely short time intervals (e.g., milliseconds), the control module sequentially illuminates the light-emitting units of different colors such as red, green, and blue, and simultaneously triggers the light receiver 3 to perform exposure and image acquisition when each color light source is illuminated. This obtains a sequence of multispectral images of the same ground scene under different color lighting conditions. Second, the control module runs image processing and dirt identification algorithms, performing in-depth analysis based on the aforementioned multispectral image data. Finally, based on the identification results, the control module outputs corresponding cleaning strategy control signals to the motion mechanism or cleaning mechanism of the autonomous cleaning equipment (such as a vacuum motor, roller brush, water tank pump, etc.), directing it to perform corresponding actions, such as focused cleaning, repeated wiping, avoidance maneuvering, or activating / deactivating specific functions.

[0032] Specifically, the main optical characteristic of solid stains 7 (such as dust, particles, and pet feces) is that they block light and form shadows. The control module's algorithm compares multiple images acquired under different wavelengths of probe light. Since the pattern and color of the ground background are fixed, their appearance (brightness) may differ under red, green, and blue light; however, due to their physical occlusion, the shadows cast by real solid stains 7 under different colors of light have highly consistent shape, position, and outline. Through image differencing, feature matching, and other techniques, these stable shadow features can be clearly extracted, effectively filtering out changing background interference. In addition, the surface of solid stains 7 also reflects some light (i.e., solid stain reflected light 8). By analyzing the intensity of this reflected light in images of different colors of light (i.e., its spectral characteristics), the color of the stain itself can be determined. For example, if an object reflects strongly under red light and weakly under blue-green light, it is likely to appear red or yellow. This helps distinguish between dark soil and light-colored paper scraps, or accurately identify brownish-yellow pet feces, preventing it from being crushed and spread.

[0033] The main optical characteristic of liquid stains 5 (such as water stains, urine, and beverages) is specular reflection. The smooth surface of the liquid produces strong directional reflection of incident light (i.e., reflected light 6 from the liquid stain). The control module locates the liquid by analyzing abnormally bright areas in multicolor light images. By analyzing the intensity distribution and color information (i.e., RGB channel values) of these bright areas in images at different wavelengths, the color and properties of the liquid can be inferred. For example, clear water strongly reflects all colors of light, so this area appears as a bright white in the three-color image. Pale yellow urine absorbs more blue light and reflects less, so its brightness in the blue light image is significantly lower than its brightness in the red and green light images. Through this spectral analysis, the system can accurately determine the type of liquid.

[0034] Furthermore, the wavelength emitted by the multi-color detection light emitter 2 in this system is not limited to visible light; it can also include invisible light bands, such as ultraviolet (UV) and / or infrared (IR). Ultraviolet light, especially the UVC band (200-280nm), has bactericidal and disinfecting effects. Infrared light, especially the mid-to-far infrared, has a thermal effect and can be used for drying. Thus, the control module can be further configured to: when specific biological contaminants (such as pet urine or vomit) are identified by visible light, in addition to planning a cleaning path, it can also control the emitter to emit ultraviolet light to irradiate the area, achieving in-situ sterilization and improving hygiene. Similarly, after cleaning (wet mopping) is completed, it can also control the emission of infrared light to dry damp areas, accelerating floor drying and preventing slippage. In this way, the dust detection system can be expanded from a purely sensing system into a comprehensive system with both sensing and processing functions.

[0035] Another aspect of the present invention provides an autonomous cleaning device, comprising a device body, a moving mechanism, a cleaning mechanism, and a control system. The control system integrates the aforementioned multi-color light-based dust detection system. Thus, by integrating the dust detection system, the autonomous cleaning device achieves visual perception-guided intelligent cleaning, realizing intelligent and precise cleaning processes.

[0036] Furthermore, such as Figure 2 As shown, the present invention also provides a dust detection method based on multicolor light, which uses the system described above for dust detection, including: The multicolor detection light emitter 2 is controlled to illuminate the ground area to be detected with a first wavelength, and the first image data is acquired through the light receiver 3; The multicolor detection light emitter 2 is controlled to switch to the second wavelength to illuminate the ground area to be detected, and the second image data is acquired through the light receiver 3; By fusing and analyzing the first and second image data, the type, color, and distribution of dirt on the ground can be identified; Based on the identification results, generate and execute corresponding cleaning strategies.

[0037] The fusion analysis includes: calculating the difference image between images of different wavelengths to eliminate static background interference; extracting feature regions from the difference image and analyzing the grayscale value or RGB component ratio of the feature region under different wavelength images to determine the color attribute of the dirt; and distinguishing between solid stains and liquid stains based on the texture, edge features, and reflective properties of the feature region. The specific analysis method is as described in the above system and will not be repeated here.

[0038] The following is a specific embodiment to illustrate the system and method provided by the present invention: Example: The autonomous cleaning device 1 is a robot vacuum cleaner that combines sweeping and mopping. A multi-color detection light emitter 2 and a light receiver 3 are installed at the front of its body. The multi-color detection light emitter 2 consists of an RGB tri-color LED module and a collimating lens, used to project a fan-shaped, essentially parallel, wide beam of light forward and downward onto the ground, covering an area approximately 0.5 meters wide and 1 meter long in front of the device. The light receiver 3 is a 2-megapixel global shutter CMOS sensor equipped with a wide-angle lens.

[0039] During operation, the control module integrated on the robot's motherboard cycles through the RGB LEDs at a frequency of 100Hz. Within a complete cycle (10ms), the R, G, and B LEDs are lit sequentially, with each LED lit for approximately 2.5ms. In the latter half of each lighting period, the CMOS sensor is precisely triggered to perform an exposure, thereby acquiring one red illumination image, one green illumination image, and one blue illumination image, respectively.

[0040] At this point, the image processing algorithm within the control module executes the following steps: S1. Denoise the three images and perform pixel-level registration to ensure that the same ground point is in the same position in the three images; S2. Calculate the difference maps between the green and red images, and between the green and blue images. In the difference maps, if the brightness of the ground background pattern is similar under different lighting conditions, the difference value is close to zero; while the stable shadow 9 formed by the solid stain 7, because it exists in different images and its position is fixed, will show significant edge features in the difference maps. Through thresholding and connected component analysis, these candidate solid stain regions can be extracted.

[0041] S3. For each candidate region, backtrack to the original R, G, and B images and read the average pixel value of that region. If the region is low in brightness (dark) in all three images, it is confirmed as a solid stain. By analyzing the relative intensity ratio of (R, G, B), its color is determined. For example, if a region is identified with a relatively high R value compared to its G and B values, it may be a red or brownish-yellow object. Combined with morphological features, it can be preliminarily judged as a pet feces risk, triggering an avoidance or gentle cleaning command.

[0042] S4. Simultaneously, the algorithm searches for regions with abnormally bright pixel values ​​across the entire image. For a candidate bright region, its brightness in the R, G, and B images is analyzed. If its brightness is high and uniform across all images, it is identified as a clear water stain; if its brightness in the B image is significantly lower than its brightness in the R and G images, it is identified as a yellowish-brown liquid (such as urine) and marked as an area requiring focused cleaning and disinfection.

[0043] Therefore, the multi-color light-based dust detection system provided by this invention utilizes multispectral information and compares image features under different colors of light to effectively distinguish between static floor background patterns and dynamic real dirt, greatly reducing the false alarm rate. It can not only detect the presence of dirt but also distinguish between solids and liquids, and further determine the color of the dirt, providing a more accurate basis for formulating cleaning strategies and achieving truly intelligent cleaning. Furthermore, this system has good detection performance on floors of different colors and materials (such as dark floors and colored carpets) and dirt of various colors, making it widely applicable.

[0044] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0045] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0046] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0047] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0048] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0049] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0050] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0051] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0052] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A multi-color light based dust visualization system, characterized in that, The dust detection system comprises: a multi-color probe light emitter (2) configured to project at least two different wavelength probe light beams (4) to a to-be-cleaned ground area in front or side-front of an autonomous cleaning device (1); a light receiver (3) configured to receive light signals from the to-be-cleaned ground area and generate corresponding image data; a control module electrically connected with the multi-color probe light emitter (2) and the light receiver (3) and configured to: control the multi-color probe light emitter (2) to switch the probe light beams (4) of different wavelengths in a time-sharing manner; based on multiple sets of image data acquired by the light receiver (3) under illumination of probe light beams of different wavelengths, perform dirt identification and output a cleaning strategy control signal to the autonomous cleaning device (1).

2. The multi-color light based dust visualization system of claim 1, wherein, The multi-color probe light emitter (2) comprises multiple independently controllable light emitting units selected from light emitting diodes (LEDs) or laser diodes (LDs), and the wavelength combination of the light emitting units includes at least two of red light, green light and blue light.

3. The multi-color light based dust visualization system of claim 2, wherein, The multi-color probe light emitter (2) further comprises an optical element configured to shape the probe light beams (4) into parallel light or linear light to cover the to-be-cleaned ground area.

4. The multi-color light based dust visualization system of claim 1, wherein, The light receiver (3) is a CMOS image sensor.

5. The multi-color light based dust visualization system of claim 1, wherein, The control module performing dirt identification comprises: for solid dirt (7), distinguishing ground background patterns from real dirt by comparing the shape, position and contour consistency of shadows (9) in images under different wavelength probe light beams, and combining the spectral characteristics of solid dirt reflection light (8) on the dirt surface to determine the color of the solid dirt (7); for liquid dirt (5), judging the nature and color of the liquid dirt (5) by analyzing the intensity distribution and color information of liquid dirt reflection light (6) in images under different wavelength probe light beams.

6. The multi-color light based dust visualization system of claim 1, wherein, The wavelengths emitted by the multi-color probe light emitter (2) further include invisible light bands, which include ultraviolet light for sterilization and / or infrared light for drying of the dirt area.

7. The multi-color light based dust visualization system of claim 6, wherein, The control module is further configured to, after identifying specific biological dirt, control the multi-color probe light emitter (2) to emit ultraviolet light for sterilization or infrared light for drying.

8. An autonomous cleaning device comprising a device body, a moving mechanism, a cleaning mechanism, and a control system, characterized in that, The control system integrates the multi-color light-based dust detection system according to any one of claims 1-7.

9. A dust visualization method based on polychromatic light, characterized in that The method uses the system according to any one of claims 1-7 to detect dust, comprising: controlling the multi-color probe light emitter (2) to irradiate a to-be-detected ground area with a first wavelength and acquiring first image data through the light receiver (3); controlling the multi-color probe light emitter (2) to switch to a second wavelength to irradiate the to-be-detected ground area and acquiring second image data through the light receiver (3); fusing and analyzing the first image data and the second image data to identify the type, color and distribution of ground dirt; generating and executing a corresponding cleaning strategy according to the identification result.

10. The multi-color light based dust visualization method of claim 9, wherein, The fusion analysis comprises: Difference images between different wavelength images are calculated to eliminate static background interference; A feature region in the difference image is extracted, and the gray value or RGB component ratio of the feature region under different wavelength images is analyzed to determine the color attribute of the stain; Based on the texture, edge feature and reflection characteristics of the feature region, solid stains and liquid stains are distinguished.

Images