JP7663129B2 - Speed estimation method, speed estimation device, speed estimation system, and computer program - Google Patents

Description

The present disclosure relates to a speed estimation method, a speed estimation device, a speed estimation system, and a computer program.

In recent years, with the increase in the number of pets and their longer lifespans, animal healthcare, which involves understanding a pet's daily health condition and amount of exercise, has been attracting attention.

Patent document 1 describes a pedometer that includes an acceleration sensor that detects the acceleration of a pedestrian, a calculation unit that calculates the acceleration data output from the acceleration sensor and outputs the calculation results, a memory unit that stores the acceleration data and the calculation results, and a display unit that displays the calculation results, and is characterized in that the calculation unit calculates the walking pitch, stride length and walking speed of the pedestrian using the period and amplitude derived from the acceleration data, and displays them on the display unit.

Patent document 2 describes an animal activity measuring device that is equipped with a three-dimensional acceleration sensor that is fixed to an animal and detects acceleration signals (Gx, Gy, Gz) in the X direction, which is the forward/backward direction of the animal, the Y direction, which is the left/right direction, and the Z direction, which is the up/down direction of the animal; a memory unit that accumulates acceleration signals from the acceleration sensor; a tilt angle calculation unit that calculates a left/right tilt angle θx about the X axis of an XY plane relative to the Z direction and a front/back tilt angle θy about the Y axis based on the acceleration signal; and a tilt angle two-dimensional display control unit that displays the left/right tilt angle θx of the XY plane and the front/back tilt angle θy of the XY plane in two-dimensional graphics on a screen.

JP 2006-118909 A JP 2011-217928 A

None of the documents describe estimating the walking speed of a quadrupedal animal, and it is desirable to understand the amount of momentum of a quadrupedal animal from its walking speed, for example. The present disclosure provides a speed estimation method, a speed estimation device, a speed estimation system, and a computer program capable of estimating the walking speed of a quadrupedal animal.

The speed estimation method disclosed herein is a speed estimation method for estimating the walking speed of a quadrupedal animal, which determines the animal's walking period using acceleration obtained from an acceleration sensor attached to the animal's body, and estimates the animal's walking speed from the walking period determined from the size of a specific part of the animal's physique and the acceleration, using a relationship between the walking period and the size and acceleration of a specific part of the animal's physique that is predefined and represents the animal's physique.

The speed estimation device disclosed herein is a speed estimation device that estimates the walking speed of a quadrupedal animal, and includes a calculation device, which determines the animal's walking period using acceleration obtained from an acceleration sensor attached to the animal's body, and estimates the animal's walking speed from the walking period determined from the size of a specific part of the animal's physique and the acceleration, using a relationship between the walking period and the size and acceleration of a specific part of the animal's physique that is predefined and the walking speed.

The speed estimation system disclosed herein is a speed estimation system for estimating the walking speed of a quadrupedal animal, and includes a measuring device having an acceleration sensor attached to the animal and detecting the animal's movement, and a communication device for transmitting the acceleration detected by the acceleration sensor, and a speed estimation device including a calculation device, and the calculation device determines the animal's walking period using the acceleration obtained from the measuring device, and estimates the animal's walking speed from the walking period determined from the size of a specific part of the animal's physique and the acceleration obtained using a relationship between the walking period and the size and acceleration of a specific part of the animal's physique that is predetermined and the walking speed.

These general and specific aspects may be realized by systems, methods, and computer programs, as well as combinations thereof.

The speed estimation method, speed estimation device, speed estimation system, and computer program disclosed herein make it possible to estimate the walking speed of a quadrupedal animal, and therefore, for example, to grasp the amount of momentum of the quadrupedal animal from its walking speed.

FIG. 1 is a schematic diagram illustrating a speed estimation system according to an embodiment. FIG. 2 is a block diagram showing a configuration of a measurement device of the speed estimation system of FIG. 1 . 2 shows an example of XYZ axes used in the velocity estimation system of FIG. 1 . 3B shows an example of the XYZ axes as viewed from a different direction from that shown in FIG. 3A. 1 is a block diagram showing a configuration of a speed estimation device according to an embodiment. 1 is a graph showing the results of measuring the walking cycle of dogs of various body lengths. 1 is a graph showing the results of measuring the walking cycle when walking at various speeds. This is a graph comparing the results of measuring the gait cycle when dogs belonging to different groups were made to walk at various speeds. 2 is an example of a waveform showing acceleration obtained by the velocity estimation system of FIG. 1 . 7 is an example of a moving average waveform obtained from the waveforms in FIG. 6. 7 is another example of a moving average waveform obtained from the waveform of FIG. 6. FIG. 7B is a diagram for explaining how to determine a walking period using the waveform in FIG. 7A. 5 is a flowchart illustrating an example of a speed estimation process in the speed estimation device of FIG. 4 . 13A and 13B are diagrams illustrating a method of determining a walking period according to the first modified example. FIG. 11 is a schematic diagram showing a speed estimation system according to a second modified example. FIG. 11 is a schematic diagram showing a speed estimation system according to a third modified example.

Below, a speed estimation system, a speed estimation device, and a speed estimation method according to the embodiments will be described with reference to the drawings. The speed estimation system, the speed estimation device, and the estimation method of the present disclosure estimate the walking speed of a quadrupedal animal. Specifically, the animal may be a quadrupedal animal that can be kept as a pet by humans. The speed estimation system according to the embodiments uses the acceleration obtained by an acceleration sensor as the motion state of the animal. Note that in the following description, the same components are denoted by the same reference symbols and description thereof will be omitted.

In the following description, unless otherwise specified, the "walking cycle" can be the period from when the right front paw lands to when the next right front paw lands. However, the "right front paw" is just one example, and can be any one of the four paws. Alternatively, the "walking cycle" can be determined using the waveform of acceleration obtained by an acceleration sensor. A method for determining the walking cycle will be described later, but for example, the walking cycle can be the period from one peak to the next peak in the waveform of acceleration measured by the acceleration sensor. The acceleration sensor used here can measure at least one of the accelerations in the x-axis, y-axis, or z-axis directions. For example, a three-axis acceleration sensor capable of measuring acceleration in three axes, x, y, and z, can be used to measure the acceleration. Alternatively, a six-axis sensor capable of measuring angular velocity in three axes, x, y, and z, can be used instead of the acceleration sensor to measure the acceleration. In this case, the walking cycle is determined using the acceleration and angular velocity.

Speed Estimation System
As shown in FIG. 1, a speed estimation system 1 according to an embodiment includes a measuring device 10 that is attached to an animal and measures acceleration that represents the state of the animal during movement. The speed estimation system 1 also includes a speed estimation device 20 that is owned by a user such as the owner of the animal. The speed estimation device 20 executes a process of receiving and registering the size of the animal, a process of determining the walking cycle of the animal from the acceleration measured by the measuring device, and a process of estimating the walking speed from the size and walking cycle of the animal. The measuring device 10 and the speed estimation device 20 are configured to be connectable via a network. In the following, an example will be described in which the animal is a dog, and the movement state of the dog during a walk is measured by an acceleration sensor, and the walking speed is estimated according to the movement state.

<Measuring Equipment>
The measurement device 10 is a device attached to a dog, which is an animal whose walking speed is to be estimated as described above. As shown in FIG. 2 as an example, the measurement device 10 includes an acceleration sensor 11, a storage device 12, a communication device 13, and a calculation device 14. The measurement device 10 can be attached to a "leash," a "collar," a "harness," or the like used for walking the dog whose walking speed is to be estimated. Here, the "leash" is a string that connects a person and a dog during a walk. The "collar" is a ring that is attached around the dog's neck. For example, the collar may be formed so that a lead can be attached to the collar. The "harness" is a so-called harness that is attached to the dog's body. The harness may be formed so that a lead can be attached to the harness. Generally, when walking a dog, one end of the lead is attached to at least one of a collar or a harness, and the other end is held by a human such as the owner.

Here, if the measuring device 10 is attached to a lead, it is possible that the movement state of the animal measured by the acceleration sensor 11 will be blurred compared to a collar or harness, since the movement caused by factors other than the dog's movement will be greater. Therefore, the measuring device 10 may be attached to a collar or harness, which is closer to the dog than a lead and less susceptible to influences other than the dog's movement. When the measuring device 10 is attached to a collar or harness, it is possible to reduce the occurrence of blurring. The measuring device may also be attached to a harness. In the case of a harness, it is possible to reduce the risk of rotation by being in close contact with the torso. Furthermore, the measuring device 10 may be attached to a position along the animal's spine. For example, the measuring device 10 may be attached to the harness so that the position of the acceleration sensor 11 in particular is positioned along the dog's spine.

The acceleration sensor 11 acquires at least one of the rate of change of the speed in the direction of travel of the animal, the left-right direction set with respect to the direction of travel, and/or the up-down direction set with respect to the direction of travel. Here, as shown in FIG. 3A and FIG. 3B, when the dog moves forward, the direction of travel is the y-axis direction, the left-right direction with respect to the y-axis which is the direction of travel is the x-axis direction, and the up-down direction with respect to the y-axis which is the direction of travel is the z-axis direction. The acceleration sensor 11 may also be a three-axis acceleration sensor capable of measuring the acceleration of three axes of the x, y and z axes. As described above, instead of the acceleration sensor 11, a six-axis sensor capable of measuring the acceleration of three axes of the x, y and z axes and the angular velocity of three axes of the x, y and z axes may be used. In addition, in order to correct the axis of the acceleration sensor 11 or the angular velocity sensor, the measurement device 10 may have a three-axis geomagnetic sensor.

The storage device 12 is a recording medium for recording various information. The storage device 12 is realized, for example, by a RAM, a ROM, a flash memory, an SSD (Solid State Drive), a hard disk drive, or other storage device, or an appropriate combination of these. The storage device 12 stores a measurement program P1, which is a computer program executed by the calculation device 14, and various data such as measurement data D1 obtained by measuring acceleration and velocity.

The communication device 13 enables data communication with an external device (e.g., the speed estimation device 20). The communication device 13 transmits the measurement data stored in the storage device 12 to the speed estimation device 20 at a predetermined timing. The predetermined timing may be, for example, regular timing during a dog walk or when the dog walk is finished. The above-mentioned data communication may be performed according to a known wireless or wired communication standard. When performing data communication while walking the dog, the communication device 13 may be wireless. The communication device 13 may employ, for example, a communication circuit capable of performing communication according to the Bluetooth (registered trademark) standard.

The arithmetic unit 14 is a controller that controls the entire measurement device 10. The arithmetic unit 14 executes the measurement program P1 stored in the memory device 12 to perform processes such as measuring acceleration using the acceleration sensor 11, storing the measurement data, and transmitting the measurement data. The arithmetic unit 14 may be various processors such as a CPU, MPU, GPU, FPGA, DSP, ASIC, etc. The arithmetic unit 14 may also be a hardware circuit designed specifically to realize a specified function.

<Speed Estimation Device>
As shown in the block diagram of Figure 4, the speed estimation device 20 includes an input device 21, an output device 22, a communication device 23, a calculation device 24, and a storage device 25, and is realized by an information processing device such as a personal computer or a mobile terminal owned by a user.

The input device 21 is an operation button, keyboard, mouse, touch panel, microphone, etc. that are used for operations and data input. The output device 22 is a display, speaker, etc. that are used for outputting processing results and data.

The communication device 23 enables data communication with the measurement device 10 and other external devices (not shown). The above-mentioned data communication can be performed according to a known communication standard. For example, the speed estimation device 20 may include, as the communication device 23, a plurality of circuits that comply with different communication standards, such as a circuit that enables wired data communication and a circuit that enables wireless data communication. For example, wired data communication is performed by using a communication controller of a semiconductor integrated circuit that operates in accordance with the Ethernet (registered trademark) standard and/or the USB (registered trademark) standard as the communication device 23. Wireless data communication is performed by using a communication controller of a semiconductor integrated circuit that operates in accordance with the IEEE 802.11 standard for LAN (Local Area Network) and/or the fourth/fifth generation mobile communication system, so-called 4G/5G, for mobile communication, as the communication device 23.

The storage device 25 is a recording medium for recording various information. The storage device 25 is realized, for example, by a RAM, a ROM, a flash memory, an SSD (Solid State Drive), a hard disk drive, or other storage device, or an appropriate combination of these. The storage device 25 stores an estimation program P2, which is a computer program executed by the calculation device 24, a corresponding model D2, size data D3, accumulated data D4, and moving average data D5.

The corresponding model D2 is a model showing the relationship between the size of a specific part of an animal's physique, the walking cycle of an animal of that size, and the walking speed of the animal. This corresponding model D2 is a model obtained by measuring the walking cycle at a specific walking speed for multiple animals whose sizes of specific parts have been measured in advance. By using the corresponding model D2, the walking speed can be estimated based on the size of a specific part of an animal and the walking cycle. For example, the corresponding model D2 may be a formula that derives the walking speed during a walk from the size of a specific part of a dog's physique and the walking cycle measured during a walk of a dog of that size. Also, for example, the corresponding model D2 may be a table that specifies the relationship between the range of the size of a specific part, the walking cycle of a dog whose specified size is in the corresponding range, and the walking speed during the walk. The size of a specific part that represents the physique can be, for example, at least one of the body length, body height, length of the forelimbs, or length of the hind limbs. Here, the "body length" of an animal refers to the length from the chest to the tip of the buttocks. The "body height" refers to the height from the ground to the back when the animal is standing on all four legs. "Forelimb length" refers to the height from the ground to the base of the front legs when the animal is standing on all fours. "Hind limb length" refers to the height from the ground to the base of the hind legs when the animal is standing on all fours.

For example, dogs with similar heights will have similar front and rear leg lengths, and similar body lengths. And dogs with similar heights will have similar walking cycles when walking at the same speed, regardless of breed. Therefore, the walking speed under the conditions can be estimated from the size of a specific part that represents the dog's physique and the walking cycle. Table 1 shows a comparison of correlation coefficients between the body length, body height, and front leg length and the walking cycle when multiple dogs are attached with acceleration sensors and made to walk on a treadmill at a speed of 3 km/h and when made to walk on a treadmill at a speed of 4 km/h. According to this, at a speed of 4 km/h, the correlation between the walking cycle and the body length, body height, and front leg length is extremely high for all of them. Also, at a speed of 3 km/h, the correlation between the body length and the walking cycle is the highest. Therefore, when estimating the walking speed from the size of a specific part that represents the physique and the walking cycle, the body length may be used as the size of a specific part that represents the physique.

Figure 5A is a graph showing the correlation between the body length and the walking cycle of a dog. In Figure 5A, the horizontal axis is the body length, and the vertical axis is the walking cycle. Line L1 in Figure 5A shows the relationship between the walking cycle and each body length measured when multiple dogs (three dogs with different body lengths) are made to wear harnesses equipped with acceleration sensors and walk on a treadmill at a speed of 3 km/h. Line L2 shows the relationship between the walking cycle and the body length measured when the same multiple dogs are made to walk on a treadmill at a speed of 4 km/h. Specifically, each line L1 and L2 in Figure 5A is a regression line obtained by plotting the relationship between the body length and walking cycle of three dogs and using the least squares method or the like based on these points. In the example of Figure 5A, the correlation coefficient between line L1 and the actual result is 0.9323, and the correlation coefficient between line L2 and the actual result is 0.9494, both of which are high values. Therefore, as shown by the straight lines L1 and L2 in FIG. 5A, the walking cycle of a dog at a given walking speed can be theoretically obtained from the sizes of given parts that represent the dog's physique.

Figure 5B is a graph showing a line L3 indicating the relationship of the walking period according to the walking speed obtained from the theoretical relationship for obtaining the walking period of the dog's walking speed from the size of a specific part of the dog as shown in Figure 5A for a certain dog. In Figure 5B, the horizontal axis is the walking speed and the vertical axis is the walking period. Also, the line L3 in Figure 5B is a regression line obtained based on the points plotted for the walking period actually measured when the dog is made to walk on a treadmill at speeds of 2 km, 3 km, 4 km, and 5 km per hour. In the example of Figure 5A, the correlation coefficient between the line L3 and the actual measurement results is also a high value of 0.9797. Therefore, in the correspondence model D2, the correspondence relationship as described above in Figures 5A and 5B is specified.

In this case, the corresponding model D2, an example of which is shown in FIG. 5A and FIG. 5B, can be a model for a dog in which at least one of the ratio of body length to body height, the ratio of body length to front leg length, or the ratio of body length to rear leg length is within a predetermined range. For example, there are multiple types of dogs, and for the majority of them, the ratio of body length to body height, and the ratio of body length to front leg or rear leg length are within a predetermined range. For example, dogs may be classified into large dogs, medium dogs, and small dogs by size, but regardless of their size, for most types of dogs, the above-mentioned ratios tend to be within a predetermined range. For dogs that meet these conditions, the relationship between walking speed and walking cycle tends to fall within a predetermined range for each size of a specific part that represents the physique, so that the walking speed can be derived from the size of a specific part that represents the physique and the walking cycle.

On the other hand, for dogs whose size of a specific body part representing the physique is outside a specific range, it is difficult to derive a walking speed corresponding to the size of the specific body part and the walking cycle using the correspondence model D2 described above with reference to Figures 5A and 5B. For example, dachshunds, corgis, etc. are so-called short-legged dog breeds, and compared to other types of dogs, the length of their legs is short relative to their body length. If the ratio of the size of the specific body part representing such a physique is different from that of most types of dogs, it is difficult to obtain a walking speed corresponding to the size of the body length, etc. and the walking cycle in the same manner as most other types of dogs. Therefore, the correspondence model D2 can be determined for each group for which a specific condition is set.

In addition to the straight line L3 shown in FIG. 5B, FIG. 5C includes a straight line L4 showing the relationship between the walking period and the walking speed of a so-called short-legged dog breed. Specifically, the straight line L3 also shown in FIG. 5B shows the relationship derived from the walking speed and walking period of a toy poodle, which was selected as an example of a general dog that is not a short-legged dog breed. In contrast, the straight line L4 shows the relationship derived from the walking speed and walking period of a dachshund, which was selected as an example of a short-legged dog breed. Specifically, the walking period actually measured when a dachshund was made to walk on a treadmill at speeds of 2.5 km/h, 4 km/h, and 4.5 km/h was plotted, and the straight line was obtained by a method similar to that of the straight line L3 based on the plotted points. In the example shown in FIG. 5C, the correlation coefficient between the straight line L4 and the actual measurement result is 0.9589, which is a high value. Therefore, in the case of a dog other than a short-legged dog breed, the walking speed is obtained using the straight line L3, and in the case of a short-legged dog breed, the walking speed is obtained using the straight line L4, so that the walking speed can be accurately obtained according to the physique of the dog.

Table 1 and FIG. 5B above show an example of using the size of a specific part to represent the animal's physique. FIG. 5C shows an example of dividing the dog into two types: so-called short-legged dog breeds, such as dachshunds, whose legs are short relative to their body length, and other dogs, such as toy poodles, whose body length to leg length ratio is normal. However, methods other than using the size of a specific part or dividing the dog into short-legged dog breeds and other than short-legged dog breeds may also be used. For example, a corresponding model D2 generated for each dog breed in more detail may be used. Also, for example, a corresponding model D2 generated for each group including several dog breeds may be used.

The size data D3 is size data of a specific part representing the physique of the dog whose walking speed is to be estimated, input via the input device 21.

The accumulated data D4 is data that accumulates the measurement data D1 received from the measuring device 10. This accumulated data D4 is data that associates acceleration and speed with the time at which they were measured. Figure 6 is an example of a waveform showing the acceleration contained in the accumulated data D4. In Figure 6, the horizontal axis represents time and the vertical axis represents sensor strength. The waveform of the accumulated data D4 shown in Figure 6 is also an example that includes all accelerations [0.1mG] on the X-axis, Y-axis, and Z-axis.

The moving average data D5 is data obtained by moving average of the waveform shown by the accumulated data D4. For example, FIG. 7A is an example of a waveform obtained by moving average of the waveform shown in FIG. 6 over three intervals. Here, "interval" refers to a certain time when the data has periodicity. In the example of FIG. 7A, one interval is 2.5 seconds. That is, the data of the first interval is obtained from 0 to 2.5 seconds, the data of the second interval is obtained from 2.5 to 5 seconds, and the data of the third interval is obtained from 5 to 7.5 seconds. The moving average of three intervals means that each is averaged as 2.5 seconds of data. Note that the 2.5 seconds of one interval is just an example, and it is ideal to divide the intervals so that the start of the first interval and the start of the second interval are in the same phase. Also, FIG. 7B is an example of a waveform obtained by moving average of the waveform shown in FIG. 6 over eight intervals.

The arithmetic unit 24 is a controller that controls the entire speed estimation device 20. The arithmetic unit 24 executes a series of processes to estimate the walking speed of the target animal by executing the estimation program P2 stored in the storage device 25. The arithmetic unit 24 may be various processors such as a CPU, MPU, GPU, FPGA, DSP, ASIC, etc. The arithmetic unit 24 may also be a hardware circuit designed specifically to realize a specified function.

<Registering size data>
The calculation device 24 acquires the sizes of the predetermined parts that represent the physique of the dog, and registers the size data D3 in the storage device 25. For example, the user inputs size data D3 indicating the sizes of the predetermined parts that represent the physique of the dog via the input device 21 at the timing of initial setup when the user starts using the speed estimation device 20. Note that for grown-up dogs, the sizes of the predetermined parts usually do not change significantly. Therefore, once the calculation device 24 registers the size data D3 input at the timing of initial setup, it can acquire, specifically read and use, this size data D3 from the storage device 25 in subsequent walking speed estimations.

Calculating walking cycle
The calculation device 24 determines the walking period of the animal using the acceleration obtained from the acceleration sensor 11 attached to the body of the animal. For example, the walking period determined by the calculation device 24 can be the period between adjacent peaks in a waveform representing the acceleration obtained by the acceleration sensor 11. Specifically, when the calculation device 24 obtains the measurement data D1 from the measurement device 10, it generates a waveform representing the acceleration and accumulates it as accumulated data D4. Furthermore, the calculation device 24 determines the walking period from the waveform represented in the accumulated data D4.

At this time, the calculation device 24 calculates the moving average of the values obtained by the acceleration sensor from the accumulated data D4, and can calculate the animal's walking period from the waveform represented by this moving average of acceleration.

8 shows an example of a walking cycle obtained by one peak (peak 1) in the waveform of FIG. 7A and a peak (peak 2) adjacent to this peak. In FIG. 8, P1 indicates the walking cycle obtained from the waveform of the Z axis. Also, P2 indicates the walking cycle obtained from the waveform of the X axis. In the example shown in FIG. 8, the peak is easy to identify in the waveform showing the acceleration in the Z axis direction, which is the vertical direction of the walking dog. Also, in the example shown in FIG. 8, the peak is easy to identify in the waveform showing the acceleration in the X axis direction, which is the horizontal direction of the walking dog. Furthermore, in the example shown in FIG. 8, the waveform showing the acceleration in the Y axis direction, which is the dog's moving direction (front-back direction), changes relatively little, but the position of each peak is linked to the waveform in the Z axis direction. In this case, the calculation device 24 can easily identify the walking cycle by using either the waveform of the acceleration showing the vertical direction or the horizontal direction. Furthermore, by further using the waveform showing the acceleration in the Y axis direction, the peak can be detected more accurately from the waveform showing the acceleration in the Z axis direction.

Note that while Figure 8 shows an example of determining a walking period from one set of adjacent vertices, the calculation device 24 can sequentially determine multiple walking periods from multiple sets of adjacent vertices. Specifically, as long as the dog continues walking, the vertices exist consecutively and the walking period is also observed consecutively. Therefore, the calculation device 24 can continuously determine each walking period using the method described above.

《Walking speed estimation》
The calculation device 24 uses the corresponding model D2 to estimate the walking speed according to the size data D3 registered in the storage device 25 and the obtained walking cycle. As described above, the calculation device 24 can continuously obtain the walking cycle, so that the calculation device 24 can estimate the walking speed continuously using the continuous walking cycle. Here, as described above with reference to FIG. 5C, when a relationship between the walking cycle and the walking speed that differs for each group is defined, the calculation device 24 estimates the walking speed using the corresponding model D2 according to the group to which the dog to be predicted belongs. For example, in the example shown in FIG. 5C, when the dog is not a short-legged dog breed, the calculation device 24 estimates the walking speed using the straight line L3. When the dog is a short-legged dog breed, the calculation device 24 estimates the walking speed using the straight line L4.

Walking speed estimation method
A walking speed estimation method implemented in the speed estimation device 20 according to the embodiment will be described with reference to the flowchart shown in Fig. 9. Since the specific processing of each step in the flowchart has been described above, each step will be described in a simplified manner.

First, at the time of registration, the calculation device 24 acquires the size of a specific part that represents the physique of the target dog via the input device 21 (S01). Here, the calculation device 24 stores the acquired size of the specific part in the storage device as size data D3.

Next, during a walk, the calculation device 24 acquires measurement data D1 including acceleration obtained by the acceleration sensor 11 from the measurement device 10 (S02). Here, the calculation device 24 stores the acquired measurement data D1 in the storage device 25 as accumulated data D4 to be accumulated.

The calculation device 24 also uses the accumulated data D4 to calculate a moving average of the acceleration (S03). Here, the calculation device 24 stores the calculated moving average in the storage device 25 as moving average data D5.

The calculation device 24 calculates the walking period using the moving average data D5 obtained in step S03 (S04). Note that although the walking period can be determined more accurately by using the moving average data D5, the walking period may also be calculated from the accumulated data D4. In that case, the process of step S03 is not necessary, and in step S04, the calculation device 24 calculates the walking period using the accumulated data D4.

Then, the calculation device 24 uses the corresponding model D2 to estimate the walking speed according to the size of the size data D3 and the walking period obtained in step S03 (S05). At this time, the calculation device 24 estimates the walking speed using the corresponding model D2 according to the group to which the dog to be predicted belongs.

The calculation device 24 also outputs the estimation result of step S05 (S06). Note that in the flowchart shown in FIG. 9, the processing of steps S02 to S06 is shown as being performed during a walk, but this is not limited to when the person is walking, and may be performed when the person finishes walking and returns home.

In this way, the speed estimation device 20 can use the acceleration obtained from the acceleration sensor while walking to determine the animal's walking period and estimate the walking speed according to the animal's size and walking period.

<Variation 1>
The example described above using FIG. 8 is an example in which the walking period is calculated using adjacent peaks of the waveform. However, the walking period is not limited to this method, and can be calculated using various methods. For example, the calculation device 24 may determine the walking period as a period including adjacent intersections among a waveform indicating acceleration obtained by the acceleration sensor 11 and an intermediate line connecting one positive peak of the waveform and an adjacent negative peak to the positive peak.

For example, as shown in FIG. 10, when an intermediate line (Middle) is provided in the middle between a peak on the positive side and a peak on the adjacent negative side, the period P3 from n1 to n5 of the intersection point between the waveform indicating acceleration and the intermediate line is set as the walking cycle. In this case, the waveform in the Z-axis direction, which is the up-down direction of a walking dog, is easy to identify the peak on the positive side and the peak on the negative side. Therefore, in this case, the calculation device 24 may obtain the walking cycle from the waveform of acceleration indicating the up-down direction. In the waveform shown as an example in FIG. 10, the period from n1 to n5 before the peak on the positive side is set as the walking cycle. However, this is not necessarily limited to this as long as it is a period including five intersection points, and for example, the period from n2 to n6 may be set as the walking cycle. In addition, FIG. 10 shows an example including five intersection points, but the number of intersection points varies depending on the shape of the waveform indicating acceleration. For example, the number of intersection points may be other numbers such as three, seven, nine, etc. In these cases, the number of intersection points is also odd. At this time, for example, the number of intersections between the Z-axis waveform and the intermediate line used in determining the walking period may be determined so as to match the walking period provisionally determined using two peaks of the X-axis waveform. The Z-axis waveform shown in FIG. 8 has a repeating pattern of large, small, large, and small peaks that are easy to understand. From the Z-axis waveform shown in FIG. 8, it is easy to accurately determine the walking period using the interval between the two peaks on the positive side. However, the difference between the size of each peak is not necessarily clearly shown as shown in FIG. 8. For example, each peak may appear at the same size. When the size of each peak is about the same, the walking period can be accurately determined by using the number of intersections between the Z-axis waveform and the intermediate line, as shown in FIG. 10.

<Modification 2>
The above-mentioned speed estimation system 1 has been described as an example including a measuring device 10 attached to an animal and a speed estimation device 20 such as a mobile terminal owned by a user, i.e., an owner. However, the speed estimation system is not limited to this form. For example, as in a speed estimation system 1A according to a second modification shown in FIG. 11, the speed estimation device 20A may be provided separately from a user terminal 30 owned by a user, i.e., an owner, and placed in the dog's home. In this speed estimation system 1A, when the dog returns home after a walk, the measurement data D1 accumulated in the storage device 12 of the measuring device 10 are transmitted together. Then, when the speed estimation device 20A acquires the measurement data D1, it executes a speed estimation process and transmits the result to the user terminal 30, and causes the user terminal 30 to output it. In this speed estimation device 20A, if the size of a specific part of the dog can be acquired from the user terminal 30 via the communication device 23, the input device 21 is not necessary.

<Modification 3>
A speed estimation system 1B according to a third modification shown in FIG. 12 is an example in which a speed estimation device 20B is present outside the dog and the owner's home in a state in which communication is possible via a network. In the speed estimation system 1B according to the third modification, the user terminal 30 transmits the measurement data D1 received from the measurement device 10 to the speed estimation device 20B. When the speed estimation device 20 acquires the measurement data D1, it executes a speed estimation process and transmits the result to the user terminal 30, which outputs the result. Note that, in the speed estimation device 20B, the input device 21 is not necessary if the size of a specific part of the dog can be acquired from the user terminal 30 via the communication device 23. For example, the speed estimation system 1B may be formed using a cloud network or the like.

<Modification 4>
In the above example, an example of predicting the walking speed using the size data D3 of the dog input by the user via the input device 21 has been described. However, the acquisition of the size data D3 of the dog is not limited to input via the input device 21. For example, dogs of the same breed generally have similar body sizes. Therefore, in the speed estimation system 1, average size data indicating the average size of a specific part of each dog breed may be stored in the storage device 25 in advance and used. In this case, when the calculation device 24 acquires the breed of the dog for which the walking speed is to be estimated input by the user via the input device 21, the calculation device 24 refers to the average size data in the storage device 25, extracts the average size of the specific part of the acquired dog breed, and predicts it as the size of the specific part of the dog. The calculation device 24 may also estimate the walking speed of the dog using the predicted size of the specific part.

<Modification 5>
Also, a size of a predetermined part of a dog may be predicted using a photo of the dog. For example, the storage device 25 stores a prediction model that has learned the relationship between a photo of the dog and the breed of the dog. The prediction model may be constructed by performing machine learning using a photo of the dog for each breed and a label of the breed of the dog as training data. Also, the storage device 25 stores average size data indicating an average size of a predetermined part of each breed. When the calculation device 24 acquires a photo of the dog, the calculation device 24 acquires a prediction result of the dog's breed from the photo of the dog using the prediction model. Also, the calculation device 24 refers to the average size data and predicts the average size of the predetermined part of the acquired breed as the size of the predetermined part of the dog. The calculation device 24 may estimate the walking speed of the dog using the predicted size.

<Modification 6>
Also, a photo of the dog may be used to predict the size of a specific part of the dog. For example, the calculation device 24 acquires a photo including a dog and a marker used to specify the size of the specific part of the dog. The marker is not limited as long as it can be used to specify the size, but for example, if the size of the measuring device 10 is fixed, the measuring device 10 may be used as the marker. Specifically, the calculation device 24 uses a photo of a dog wearing a harness to which the measuring device 10 is attached, and can predict the size of the specific part of the dog from the relationship between the size ratio of the dog included in the photo and the measuring device 10 as a marker. Also, the calculation device 24 may estimate the walking speed of the dog using the predicted size.

<1> A speed estimation method for estimating a walking speed of a quadrupedal animal, comprising:
determining a walking period of the animal using acceleration obtained from an acceleration sensor attached to the body of the animal;
estimating the walking speed of the animal from the acquired size of a predetermined part representing the animal's physique and the walking period calculated from the acceleration, using a predetermined relationship between the size of a predetermined part representing the animal's physique, the walking period, and the walking speed;
Speed estimation method.
<2> The animal is a dog,
The size is at least one of body length, body height, and forelimb length.
The speed estimation method according to <1>.
<3> The animal is a dog whose body length to body height ratio or body length to front leg length ratio is within a predetermined range.
The speed estimation method according to <2>.
<4> The walking cycle is a section between one peak and an adjacent peak of a waveform representing acceleration obtained by the acceleration sensor.
The speed estimation method according to any one of <1> to <3>.
<5> The waveform showing the acceleration is a moving average of values obtained by an acceleration sensor.
The speed estimation method according to <4>.
<6> acquiring the breed of the dog whose walking speed is to be estimated;
referring to average size data showing the average size of the predetermined part for each dog breed, predicting the average size of the predetermined part of the dog breed as the size of the predetermined part of the dog;
Using the predicted size, estimate the walking speed of the dog.
The speed estimation method according to <2>.
<7> Acquire the size of the predetermined part of the animal,
The obtained size is used to estimate the walking speed of the animal.
The speed estimation method according to any one of <1> to <5>.
<8> Obtaining a photo of the dog,
predicting the breed of the dog from the captured photograph using a predictive model that predicts the breed of the dog from a photograph of the dog;
referring to average size data showing the average size of the predetermined part for each dog breed, predicting the average size of the predetermined part of the dog breed as the size of the predetermined part of the dog;
Using the predicted size, estimate the walking speed of the dog.
The speed estimation method according to <2>.
<9> Acquire a photograph including the animal and a marker used to specify the size of the predetermined part of the animal;
predicting a size of the predetermined part of the animal based on a relationship between the animal and the marker included in the photograph;
The predicted size is used to estimate the average speed of the animal.
The speed estimation method according to any one of <1> to <5>.
<10> Acquiring the size of the predetermined part of the animal by either input by a user via an input device or reading from a storage device;
The obtained size is used to estimate the walking speed of the animal.
The speed estimation method according to any one of <1> to <5>.
<11> A speed estimation device for estimating a walking speed of a quadrupedal animal, comprising: a calculation device;
The computing device includes:
determining a walking period of the animal using acceleration obtained from an acceleration sensor attached to the body of the animal;
estimating the walking speed of the animal from the acquired size of a predetermined part representing the animal's physique and the walking period calculated from the acceleration, using a predetermined relationship between the size of a predetermined part representing the animal's physique, the walking period, and the walking speed;
Speed estimator.
<12> A speed estimation system for estimating a walking speed of a quadruped animal, comprising:
a measuring device including an acceleration sensor attached to the animal and detecting the movement of the animal, and a communication device transmitting the acceleration detected by the acceleration sensor;
a speed estimation device including a calculation device,
The computing device includes:
Using the acceleration acquired from the measurement device, a walking period of the animal is calculated;
estimating the walking speed of the animal from the acquired size of a predetermined part representing the animal's physique and the walking period calculated from the acceleration, using a predetermined relationship between the size of a predetermined part representing the animal's physique, the walking period, and the walking speed;
Speed estimation system.
<13> The measuring device is attached to a position along the spine of the animal.
The speed estimation system according to <12>.
<14> The acceleration sensor acquires at least one of a rate of change in the moving direction of the animal, a left-right direction set with respect to the moving direction, and/or a rate of change in an up-down direction set with respect to the moving direction.
The speed estimation system according to <12> or <13>.
<15> A computer program for causing a calculation device to execute the speed estimation method according to any one of <1> to <10>.
As described above, examples of the technology disclosed in the present application have been shown and the above embodiments have been described. However, the technology in the present disclosure is not limited to these, and may be applied to embodiments in which modifications, substitutions, additions, omissions, etc. are made as appropriate.

The speed estimation method, speed estimation device, and speed estimation system described in all claims of the present disclosure are realized by cooperation with hardware resources (e.g., processor, memory) and computer programs.

The speed estimation method, speed estimation device, speed estimation system, and computer program disclosed herein are useful for understanding the walking speed of four-legged animals.

Reference Signs List 1, 1A, 1B Speed estimation system 10 Measurement device 20, 20A, 20B Speed estimation device 30 User terminal 11 Acceleration sensor 12 Storage device 13 Communication device 14 Calculation device 21 Input device 22 Output device 23 Communication device 24 Calculation device 25 Storage device D1 Measurement data D2 Corresponding model D3 Size data D4 Accumulated data D5 Moving average data P1 Measurement program P2 Estimation program

Claims (15)

A speed estimation method for estimating a walking speed of a quadrupedal animal, comprising the steps of:
determining a walking period of the animal using acceleration obtained from an acceleration sensor attached to the body of the animal;
estimating the walking speed of the animal from the acquired size of a predetermined part representing the animal's physique and the walking period calculated from the acceleration, using a predetermined relationship between the size of a predetermined part representing the animal's physique, the walking period, and the walking speed;
Speed estimation method. The animal is a dog,
The speed estimation method according to claim 1 , wherein the size is at least one of a body length, a body height, and a length of a forelimb. The speed estimation method according to claim 2 , wherein the animal is a dog having a ratio of body length to body height or a ratio of body length to front leg length within a predetermined range. The speed estimation method according to claim 1 , wherein the walking period is a section between one peak and an adjacent peak of a waveform representing the acceleration obtained by the acceleration sensor. The speed estimation method according to claim 4 , wherein the waveform indicating the acceleration is a moving average of values obtained by an acceleration sensor. obtaining a breed of the dog whose walking speed is to be estimated;
referring to average size data showing the average size of the predetermined part for each dog breed, predicting the average size of the predetermined part of the dog breed as the size of the predetermined part of the dog;
The method of claim 2 , further comprising estimating a walking speed of the dog using the predicted size. Obtaining the size of the predetermined part of the animal;
The obtained size is used to estimate the walking speed of the animal.
The method of claim 1 . obtaining a photograph of the dog;
predicting the breed of the dog from the captured photograph using a predictive model that predicts the breed of the dog from a photograph of the dog;
referring to average size data showing the average size of the predetermined part for each dog breed, predicting the average size of the predetermined part of the dog breed as the size of the predetermined part of the dog;
The method of claim 2 , further comprising estimating a walking speed of the dog using the predicted size. acquiring a photograph including the animal and a marker that is used to identify the size of the predetermined part of the animal;
predicting a size of the predetermined part of the animal based on a relationship between the animal and the marker included in the photograph;
The predicted size is used to estimate the average speed of the animal.
The method of claim 1 . acquiring a size of the predetermined part of the animal by either inputting the size by a user via an input device or reading the size from a storage device;
The obtained size is used to estimate the walking speed of the animal.
The method of claim 1 . A speed estimation device for estimating a walking speed of a quadruped animal, comprising: a calculation device;
The computing device includes:
determining a walking period of the animal using acceleration obtained from an acceleration sensor attached to the body of the animal;
estimating the walking speed of the animal from the acquired size of a predetermined part representing the animal's physique and the walking period calculated from the acceleration, using a predetermined relationship between the size of a predetermined part representing the animal's physique, the walking period, and the walking speed;
Speed estimator. A speed estimation system for estimating a walking speed of a quadruped animal, comprising:
a measuring device including an acceleration sensor attached to the animal and detecting the movement of the animal, and a communication device transmitting the acceleration detected by the acceleration sensor;
a speed estimation device including a calculation device,
The computing device includes:
Using the acceleration acquired from the measurement device, a walking period of the animal is calculated;
estimating the walking speed of the animal from the acquired size of a predetermined part representing the animal's physique and the walking period calculated from the acceleration, using a predetermined relationship between the size of a predetermined part representing the animal's physique, the walking period, and the walking speed;
Speed estimation system. The speed estimation system of claim 12 , wherein the measuring device is attached to a position along the animal's spine. The speed estimation system according to claim 12 or 13, wherein the acceleration sensor acquires at least one of a rate of change in a moving direction of the animal, a left-right direction set with respect to the moving direction, and/or a rate of change in an up-down direction set with respect to the moving direction. A computer program product for causing a computing device to execute the speed estimation method according to claim 1 .

Images