JP6969354B2 - Self-propelled vacuum cleaner - Google Patents

Description

The present invention relates to a self-propelled vacuum cleaner that self-propells and cleans.

In order to drive a self-propelled vacuum cleaner for home use, the self-propelled vacuum cleaner is equipped with obstacle detection sensors such as bumper sensors and ultrasonic sensors, and the direction of travel is achieved by finding or contacting obstacles. It is known to randomly change such things as running, for example, prepare some patterns to change the direction etc., and run by combining them, or make a map of the room while running. It is automatically generated or run regularly.

On the other hand, for dust collection, a method of operating a cylindrical rotating brush to collect dust is common in order to efficiently guide dust and sand on the floor surface to the dust suction port.

Cleaning according to the type of floor will be described here. Floors are roughly divided into three types: long-haired carpets, flooring, and tatami mats with soft and easily damaged surfaces. The idea of changing the above has been proposed for a long time, and is described in, for example, Patent Document 1.

Specifically, in the case of carpet cleaning, particularly long-haired carpets tend to get entangled with the rotating brush, that is, the rotating brush is difficult to rotate (the current value of the motor that drives the rotating brush increases) and is controlled. The device slows down the rotation speed of the rotating brush so that the carpet hair does not get entangled, while increasing the power of the suction motor to compensate for the decrease in dust collection power of the rotating brush (this carpet detection operation is cleaned according to the type of floor surface). The overwhelming majority of manufacturers list that it is possible).

In the case of flooring, dust tends to collect in the part (groove) that connects the boards, so set the rotation speed high so that it can be scraped off firmly with a brush.

In the case of tatami mats, it is desirable to reduce the number of rotations and operate the rotating brush along the direction of the tatami joints (joint direction) because the material is soft and the brush operation at high rotation may damage the tatami mats.

Here, the detection of the joint direction of tatami mats by the floor sensor of this type of conventional self-propelled vacuum cleaner will be described with reference to FIGS. 5 to 7.

5 is a structural diagram of a tatami mat, FIG. 6 is a camera image of a floor sensor mounted on a conventional self-propelled vacuum cleaner, and FIG. 7 is a black-and-white converted image of the camera image.

In FIGS. 5 to 7, the tatami mat is made into a rug by weaving the warp threads 90a to 90f made of cotton or linen with the threads 91a to 91b made of rush, and the weave 92a and the next weave 92b are formed. The space is called the tatami joint (tatami joint) 93, and the direction is called the joint direction 94.

A floor sensor having a camera is attached to the lower part (bottom surface) of the main body of the conventional self-propelled vacuum cleaner, and the floor sensor acquires a camera image by the camera. The camera image shown in FIG. 6 shows at least two joints 93, and the joint direction 94 can be easily determined by a person.

FIG. 7 shows a camera image converted into a black-and-white image by a calculation means. In this kind of joint direction detection, a black-and-white image or a gray scale is converted into a black-and-white image and a portion having a strong black-and-white contrast is a feature point. It is extracted and the feature points are connected to form a line segment, and the direction of this line is detected as the joint direction.

Japanese Unexamined Patent Publication No. 2007-319485

However, with the above-mentioned conventional self-propelled vacuum cleaner, it is not easy to correctly detect the joint direction of tatami mats.

Specifically, the tatami mats appearing in past patent documents do not have the concept of weaving straw grass into the warp threads 90a to 90f, that is, the tatami mats are premised on the state in which long threads are laid in parallel, and any part of the tatami mats is used. Even if the joint direction is detected from the extracted camera image, it is described as if it is in a fixed direction.

On the other hand, since the actual tatami mat weaves straw grass at the boundaries of the warp threads 90a to 90f as described above, the boundary portion 95 is particularly characterized by large irregularities and shadows, that is, the contrast of the boundary portion is strong in the camera image. The line segment 96 connecting the boundary portions was erroneously detected as the joint direction, and there was a problem that the tatami mat surface was damaged by operating the rotating brush in the wrong joint direction.

The present invention solves the above-mentioned conventional problems, and is a self-propelled type capable of improving the detection accuracy of the joint direction of tatami mats using a floor sensor and reducing damage caused by a rotating brush when cleaning tatami mats. The purpose is to provide a vacuum cleaner.

In order to solve the above problems, the self-propelled vacuum cleaner of the present invention is a self-propelled vacuum cleaner that cleans while moving the floor surface, and uses a camera attached to the lower part of the vacuum cleaner body to make a joint on the floor surface. It is equipped with a floor sensor that detects the direction and a control device that controls the vacuum cleaner body. The floor sensor divides the image captured by the camera into a plurality of blocks and detects the joint direction for each block. However, among the joint directions detected for each block, the joint direction with a high frequency of occurrence is adopted, and the control device is characterized in that the vacuum cleaner main body is driven based on the joint direction detected by the floor surface sensor. ..

The self-propelled vacuum cleaner of the present invention accurately detects the joint direction of the tatami mat without being affected by the unevenness of the weaving boundary portion of the tatami mat surface, and operates the rotating brush along the joint direction to the tatami mat. It is possible to provide a self-propelled vacuum cleaner capable of collecting dust with less damage.

Schematic cross-sectional view of the self-propelled vacuum cleaner according to the first embodiment of the present invention. Functional block diagram of the self-propelled vacuum cleaner A diagram showing images of joints detected for each block of the floor sensor of the self-propelled vacuum cleaner. Occurrence frequency graph of joint direction (frequency) detected for each block of the same floor sensor Structural drawing of tatami mat A diagram showing a camera image of a floor sensor mounted on a conventional self-propelled vacuum cleaner. A diagram showing a black-and-white converted image of the same camera image

The first invention is a self-propelled vacuum cleaner that cleans while moving the floor surface, a floor surface sensor that detects the joint direction of the floor surface by using a camera attached to the lower part of the vacuum cleaner main body, and the cleaning. The floor sensor includes a control device for controlling the main body of the machine, the floor sensor divides the image captured by the camera into a plurality of blocks and detects the joint direction for each block, and the control device is the floor sensor. When the surface to be cleaned is, for example, a tatami mat, the camera image of the tatami mat is divided into several blocks and the joint direction is detected for each block. By using the joint direction data for each block, the joint direction of the tatami mat is accurately detected without being affected by the unevenness of the weaving boundary part of the tatami mat surface, and the vacuum cleaner body is run along the joint direction of the tatami mat. Because it can be done, the damage to the tatami can be reduced.

Further, in the second invention, in particular, the floor sensor of the first invention adopts the joint direction having a high frequency of occurrence among the joint directions detected for each block, and is affected by the unevenness of the woven boundary portion of the tatami mat surface. Not only that, even when the joint direction of each block cannot be detected due to partial scratches or dirt on the tatami mat surface, the joint direction of the tatami mat can be detected accurately without being affected by these.

In a third aspect of the invention, in particular, the self-propelled vacuum cleaner of the first or second invention comprises a rotating brush for collecting dust, and the control device is described above along the joint direction detected by the floor surface sensor. The rotating brush is operated, and since the rotating brush is operated along the joint direction, it is possible to collect dust with less damage to the tatami mat.

In the fourth invention, in particular, the control device of any one of the first to third inventions detects the type of the floor surface by the pattern in the joint direction detected by the floor surface sensor for each block, and is conventionally distinguished. It is possible to accurately detect flooring and tatami mats, which are the types of floors that are difficult to stick to.

In the fifth invention, in particular, the control device of the third or fourth invention controls the number of rotations of the rotating brush according to the type of the floor surface. For example, when the floor surface is flooring, dust is a plate. Since it tends to collect in the part (groove) that connects each other, increase the rotation speed of the rotating brush to scoop out dust firmly, while if the floor is tatami, reduce the rotation speed to reduce damage to the tatami mat. can do.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. The present invention is not limited to this embodiment.

(Embodiment 1)
FIG. 1 is a schematic cross-sectional view of the self-propelled vacuum cleaner according to the first embodiment of the present invention, FIG. 2 is a functional block diagram of the self-propelled vacuum cleaner, and FIG. 3 is a self-propelled vacuum cleaner. An image in which joints are detected for each block of the floor sensor, FIG. 4 is a graph of the occurrence frequency of joint directions (frequency) detected for each block of the floor sensor.

The configuration of the self-propelled vacuum cleaner in the present embodiment is the same as that of the conventional self-propelled vacuum cleaner, except that the method of detecting (calculating) the joint direction of the tatami mat by the floor sensor is different.

In FIGS. 1 and 2, the self-propelled vacuum cleaner 100 according to the present embodiment includes a floor surface sensor 10 provided on the lower portion (lower surface) of the vacuum cleaner main body 100a, a rotary brush 14 for scraping up dust on the floor surface, and the like. A dust collecting motor 15 that generates a suction air for sucking up the dust, a dust box 16 that collects the sucked dust, drive wheels 17a and 17b for running the vacuum cleaner body 100a, and a floor. A control device 13 for controlling a rotary brush 14, a dust collecting motor 15, and drive wheels 17a and 17b according to an output signal of the surface sensor 10 is provided.

The floor sensor 10 is composed of a camera 11 and a calculation means 12, and the camera 11 is to a personal computer (not shown) that is generally sold except that the lens portion has a slightly wide angle and can shoot a wide floor surface. There is no big difference in performance from a WEB camera with about 1.2 million pixels connected via USB.

The image acquired by using this camera 11 shows at least two joints 93, and as shown in FIG. 6, the camera 11 is attached at a position where about four joints can be photographed in the present embodiment. ing.

The camera image obtained in this way is converted into a black-and-white image by the arithmetic means 12, and further divided into, for example, 80 blocks 11a having a length shorter than the length of the joint 93 as one side as shown in FIG. A portion with strong black-and-white shading (contrast) is extracted as a feature point for each block 11a, and a line segment is formed by connecting the feature points, and the direction of this line is set as the joint direction 97a to 97z for each block 11a. , 0 degree to 359 degree is detected as a 1 degree unit orientation.

As shown in FIG. 4, when the number of joint directions 97a to 97z detected for each block 11a is counted, two joint directions, which are extreme directions of the occurrence frequency 1 group 98 and the occurrence frequency 2 group 99, are extracted. NS. The occurrence frequency 1 group 98 is mainly due to the joint direction 94 to be detected, and the occurrence frequency 2 group 99 is mainly due to the line segment generated from the boundary portion 95 of the weave, that is, the incorrect joint direction.

Here, since the length of one side of the block 11a was set to be shorter than the length of the joint 93 when the camera image was divided into blocks, the number of occurrences of the occurrence frequency 1 group 98 is larger than the number of occurrences of the occurrence frequency 2 group 99. Is overwhelmingly large.

Returning to the story, the calculation means 12 detects the direction having the largest number of occurrences as the joint direction of the entire tatami mat table. In the case of FIG. 4, it is 89 degrees, and of course, it is included in the occurrence frequency 1 group 98.

As a result, the camera image of the tatami mat is divided into several blocks 11a, the joint direction is detected for each block 11a, and the joint direction data for each of these blocks 11a is used to form the weaving boundary portion 95 of the tatami mat. Not only is it not affected by the unevenness of the tatami mat, but even when the joint direction of each block 11a cannot be detected due to partial scratches or dirt on the tatami mat surface, the joint direction of the tatami mat can be detected accurately without being affected by these.

When the floor sensor 10 accurately detects the joint direction of the tatami mat surface in this way, the control device 13 rotates the rotary brush 14 along the joint.

As a result, by operating the rotating brush 14 along the joint direction and running the vacuum cleaner main body 100a, it is possible to collect dust with less damage to the tatami mats.

Further, the calculation means 12 can detect the type of the floor surface not only as the joint direction of the entire tatami mat surface but also by the pattern of the joint direction detected for each block. For example, in the joint directions 97c, 97d, 97e detected for each block 11a as shown in FIG. 1, after the horizontal joints appear continuously, the vertical joint direction 97f appears, and then the horizontal direction again. 97g and 97h appear in succession in the joint direction. When such a pattern is detected, it is detected that the type of floor surface is tatami mat. At the same time, when the calculation means 12 determines that the type of the floor surface is tatami mat, the control device 13 reduces the rotation speed of the rotating brush 14.

In the self-propelled vacuum cleaner of the present embodiment configured as described above, it is possible to accurately detect flooring and tatami mats, which are types of flooring that are difficult to distinguish in the past, and dust is collected between plates in the flooring. Since it is easy to collect in the connecting part (groove), the rotation speed can be set high because it is firmly scraped out with the rotating brush 14, while the damage to the tatami can be reduced by reducing the rotation speed in the tatami mat.

As described above, the self-propelled vacuum cleaner according to the present invention can accurately detect the joint direction of tatami mats without being affected by the unevenness of the weaving boundary portion of the tatami mat surface. It can also be applied to self-propelled transport vehicles that transport luggage in warehouses.

10 Floor sensor 11 Camera 11a Block 12 Calculation means 13 Control device 14 Rotating brush 15 Dust collection motor 16 Dust box 17a, 17b Drive wheels 90a to 90f Warp 91a, 91b Thread 92a, 92b Weave 93 Joint 94, 97a to 97z Joint direction 95 Boundary part 96 Lines 98 Occurrence frequency 1 group 99 Occurrence frequency 2 groups 100 Self-propelled vacuum cleaner 100a Vacuum cleaner body

Claims (4)

In a self-propelled vacuum cleaner that cleans while moving the floor
A floor sensor that detects the joint direction of the floor using a camera attached to the bottom of the vacuum cleaner body,
A control device for controlling the vacuum cleaner body is provided.
The floor sensor divides the image captured by the camera into a plurality of blocks, detects the joint direction for each block, and adopts the joint direction that occurs most frequently among the joint directions detected for each block. death,
The control device is a self-propelled vacuum cleaner characterized in that the vacuum cleaner main body is driven based on the joint direction detected by the floor surface sensor. The self-propelled vacuum cleaner according to claim 1 , further comprising a rotating brush for collecting dust, wherein the control device operates the rotating brush along a joint direction detected by the floor sensor. .. The self-propelled vacuum cleaner according to claim 1, further comprising a rotating brush for collecting dust, wherein the control device controls the number of rotations of the rotating brush according to the type of the floor surface. Wherein the control device, the self-propelled cleaner according to claim 1, wherein the floor sensor and detecting the type of the floor surface by joint direction pattern detected for each block ..

Images