Sensor, method for manufacturing a sensor, and method for processing information

Abstract

To provide a high precision sensor having a simpler structure.SOLUTION: According to one aspect of the present invention, a sensor for measuring objects is provided. The sensor includes a plurality of light-emitting elements and at least one light-receiving element. The light-emitting elements are each provided at a different position on a substrate. The light-receiving element is provided on the substrate. The light-receiving element receives, as main light, one reflected light from among reflected light attributed to each light-emitting element, and receives, as crosstalk light, reflected light other than the main light in a manner such that the crosstalk light can be distinguished from the main light. The reflected light is emitted from each of the light-emitting elements and reflected from an object. On the basis of the crosstalk light and main light that are received in a distinguishable manner, spatial physical quantities related to a reference plane of the sensor and the object are measured.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] The present invention relates to a sensor, a method for manufacturing the sensor, and an information processing method.

Background Art

[0002] Patent Document 1 discloses a sensor. This sensor includes a base material, a light emitting element, a light receiving element, and a synchronous detection circuit. The light emitting elements are provided in a plurality of different positions on the base material and are configured to be able to emit light using first and second modulation signals orthogonal to each other. The light receiving element is disposed inside the base material and is configured to receive external light through a pinhole and generate a photocurrent. The synchronous detection circuit is configured to be able to detect the phase of a specific component of the photocurrent. Here, the specific component is a component caused by the combined reflected light irradiated from the light emitting element and reflected by the object, and the phase is represented as a function of the distance to the object.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] On the other hand, different from the prior art disclosed in Patent Document 1, there is a need for a sensor that can grasp spatial physical quantities such as distance with a simpler structure and high precision.

[0005] In view of the above circumstances, the present invention aims to provide a sensor having a simpler structure and high precision.

Means for Solving the Problems

[0006] According to one aspect of the present invention, a sensor for measuring an object is provided. This sensor comprises a plurality of light-emitting elements and at least one light-receiving element. The light-emitting elements are each provided at different positions on a substrate. The light-receiving element is provided on the substrate. One of the reflected light originating from each light-emitting element is received as the principal light, and the other reflected light is received as crosstalk light in a manner that distinguishes it from the principal light. The reflected light is the light emitted from each light-emitting element and reflected from the object, respectively. Based on the principal light and the crosstalk light that are received in a manner that distinguishes it, a spatial physical quantity relating to the reference surface and the object in the sensor is measured.

[0007] According to this disclosure, a sensor with a simpler structure and high accuracy can be realized. [Brief explanation of the drawing]

[0008] [Figure 1] This is a plan view showing the configuration overview of the detection unit 2 of the sensor 1 according to this embodiment. [Figure 2] This block diagram shows the configuration overview of the information processing unit 4, which is the sensor 1 according to this embodiment. [Figure 3] This block diagram shows the functions implemented by the processor 43 and other components in sensor 1. [Figure 4] This is a schematic diagram showing the neural network configuration in the pre-trained model 5. [Figure 5] This is an activity diagram showing the flow of the measurement process for object Ob using sensor 1. [Figure 6] This is a schematic diagram showing the timing of light emission from each light-emitting element 31 and the manner in which light is received by the light-receiving element 32, with the light-emitting elements 31 emitting light in the order of Figure 6A, Figure 6B, Figure 6C, and Figure 6D. [Figure 7] This is a schematic front view showing the configuration of sensor 1 in a modified example. [Modes for carrying out the invention]

[0009] [Embodiment] Embodiments of the present invention will be described below with reference to the drawings. The various features shown in the embodiments below can be combined with each other.

[0010] Incidentally, the program for implementing the software appearing in this embodiment may be provided as a non-transitory computer-readable medium, or it may be provided so that it can be downloaded from an external server, or it may be provided so that the program is launched on an external computer and its functions are realized on a client terminal (so-called cloud computing).

[0011] Furthermore, in this embodiment, "part" may include, for example, hardware resources implemented by circuits in a broad sense, and the information processing of software that can be specifically realized by these hardware resources. In addition, various types of information are handled in this embodiment, and these types of information can be represented, for example, by the physical values ​​of signal values ​​representing voltage and current, the high or low values ​​of signal values ​​as a set of binary bits composed of 0s or 1s, or by quantum superposition (so-called qubits), and communication and calculations can be performed on circuits in a broad sense.

[0012] Furthermore, a circuit in a broad sense is a circuit realized by combining at least a suitable combination of circuits, circuits, processors, and memory. In other words, it includes application-specific integrated circuits (ASICs), programmable logic devices (for example, simple programmable logic devices (SPLDs), complex programmable logic devices (CPLDs), and field programmable gate arrays (FPGAs)), etc.

[0013] 1. Hardware Configuration This section describes the hardware configuration of sensor 1 according to this embodiment. Sensor 1 is a sensor for measuring an object Ob. More specifically, sensor 1 is configured to measure spatial physical quantities relating to a reference plane in sensor 1 and the object Ob. Here, the spatial physical quantities preferably include at least the distance d between the reference plane and the object Ob, and the angles θ and φ of the object Ob with respect to the reference plane (parameters indicating the three-dimensional orientation). Particularly preferred is that the distance d is very close, for example, 50 mm or less, specifically 50, 49, 48, 47, 46, 45, 44, 43, 42, 41, 40, 39, 38, 37, 36, 35, 34, 33, 32, 31, 30, 29, 28, 27, 26, 25, 24, 23, 22, 21, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0 mm, and may also be within the range of any two of the values ​​exemplified here. According to this embodiment, the distance d and angles θ, φ with respect to the object Ob can be measured and applied to various situations such as object picking and inspection. Furthermore, according to this embodiment, the object Ob located at a very close distance can be accurately identified.

[0014] Figure 1 is a plan view showing the configuration overview of the detection unit 2 of the sensor 1 according to this embodiment. Figure 2 is a block diagram showing the configuration overview of the information processing unit 4 of the sensor 1 according to this embodiment. As shown in Figures 1 and 2, the sensor 1 mainly comprises a detection unit 2 and an information processing unit 4.

[0015] 1.1 Detection Unit 2 First, let's describe the detection unit 2. As shown in Figures 1 and 2, the detection unit 2 comprises a base material 21, a light-receiving block 3 provided on the base material 21, and an AI input unit 22 connected to the information processing unit 4. The following describes each component in more detail.

[0016] (Base material 21) The base material 21 constitutes the housing of the sensor 1. In the present embodiment, the base material 21 has, for example, a substantially circular shape and is configured in a flat plate shape, but this is merely an example and is not limited thereto. Preferably, the base material 21 is formed of a material that is, for example, black, has no gloss, and has a low reflectance. Further, the base material 21 includes a light emitting and receiving block 3 provided thereon.

[0017] (Light emitting and receiving block 3) As shown in FIG. 1, a plurality of light emitting and receiving blocks 3 are provided on the base material 21. In each light emitting and receiving block 3, a light emitting element 31 and a light receiving element 32 are arranged. Here, in one light emitting and receiving block 3, one light emitting element 31 and one light receiving element 32 are arranged. In other words, the light receiving element 32 and the associated light emitting element 31 form the light emitting and receiving block 3 adjacent to each other. A plurality of light emitting and receiving blocks 3 are provided on the base material 21.

[0018] Here, preferably, the number of light emitting elements 31 is 2 to 30, more preferably 2 to 10, and most preferably the number of light emitting elements 31 is 2 to 4. Specifically, for example, the number of light emitting elements 31 is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, and may be within the range between any two of the numerical values exemplified here.

[0019] Also, the number of the light emitting and receiving blocks 3 is preferably 2 or more, more preferably 3 or more. In this embodiment, four light emitting and receiving blocks 3a, 3b, 3c, and 3d are provided. As shown in FIG. 1, more preferably, each of the light emitting and receiving blocks 3 is arranged on the base material 21 in an annular shape and at equal intervals. In other words, the light emitting and receiving blocks 3 are arranged symmetrically with respect to the base material 21. Specifically, a light emitting element 31a and a light receiving element 32a are arranged in the light emitting and receiving block 3a, a light emitting element 31b and a light receiving element 32b are arranged in the light emitting and receiving block 3b, a light emitting element 31c and a light receiving element 32c are arranged in the light emitting and receiving block 3c, and a light emitting element 31d and a light receiving element 32d are arranged in the light emitting and receiving block 3d. According to such an aspect, symmetry occurs in each of the photocurrents J (an example of an electrophysical quantity) based on the plurality of light receiving elements 32, and the accuracy of the sensor 1 can be further improved. In other words, the number of the light receiving elements 32 is the same as that of the light emitting elements 31, and each of the light receiving elements 32 is associated with one of the light emitting elements 31 in a one-to-one manner. According to such an aspect, since the photocurrents J based on the plurality of light receiving elements 32 can be obtained, the accuracy of the sensor 1 can be further improved. The photocurrent J will be described later.

[0020] As shown in FIG. 1, the light emitting elements 31 are provided at different positions on the base material 21. The light emitting element 31 may be an element that irradiates diffused light. For example, it is a light emitting diode (LED), and preferably, it is an infrared LED that emits infrared light that is invisible to humans and harmless to the human body. Of course, it is not limited to this, and the light emitting element 31 may be a red LED, a green LED, or a blue LED. By connecting the positive side of a power source (not shown) to the anode side of the light emitting element 31 which is such a light emitting diode, a current flows and diffused light of a specific frequency is irradiated onto the object Ob.

[0021] As shown in Figure 1, the light-receiving element 32 is provided on the substrate 21. The light-receiving element 32 is an element that detects received light, and this triggers the generation of a photocurrent J, which is an example of an electrical physical quantity. In other words, the light-receiving element 32 is configured to output a photocurrent J according to the illuminance of the received light. Preferably, the relationship between the illuminance of the light and the photocurrent J is linear. Examples of main light-receiving elements 32 include photocells, photomultiplier tubes, phototransistors utilizing the internal photoelectric effect of semiconductors, photodiodes, avalanche photodiodes, photoconductive cells, and image sensors. Preferably, the light-receiving element 32 is a photodiode with broad directivity or a phototransistor that combines the photodiode with an amplifier.

[0022] Here, the light-receiving element 32 receives the reflected light L emitted from the light-emitting element 31 belonging to the same light-receiving block 3 and reflected from the object Ob as the primary light. The light-receiving element 32 also receives the reflected light L emitted from the light-emitting element 31 belonging to the adjacent light-receiving block 3 and reflected from the object Ob as crosstalk light. For example, the light-receiving element 32a receives the reflected light L originating from the light-emitting element 31a as the primary light, and receives the reflected light L originating from the light-emitting elements 31b and 31d as crosstalk light. In other words, the light-receiving element 32 is configured to receive one of the reflected light L originating from each light-emitting element 31 as the primary light, and to receive the other reflected light L as crosstalk light in a way that is distinguishable from the primary light, and to generate a photocurrent J (an example of an electrical physical quantity) corresponding to the reflected light L. The reflected light L here is the light emitted from each light-emitting element 31 and reflected from the object Ob, respectively. In this way, spatial physical quantities relating to the reference plane and the object Ob in sensor 1 are measured based on the photocurrent J, which is derived from the main light that is received in a distinguishable manner and the crosstalk light. This will be described in more detail later.

[0023] (AI input section 22) The AI ​​input unit 22 inputs the value of the photocurrent J or a calculated value based on the photocurrent J as an input parameter to the trained model 5 stored in the memory unit 42 of the information processing unit 4, which will be described later. In other words, the AI ​​input unit 22 may be a conductor leading from each photodetector 32 in the detection unit 2 to the information processing unit 4. The photocurrent J output from the photodetector 32 is input to the trained model 5 stored in the memory unit 42 via the AI ​​input unit 22 and the communication unit 41, which will be described later.

[0024] 1.2 Information Processing Unit 4 Next, the information processing unit 4 will be described. The information processing unit 4 is a processing circuit configured to control the operation of sensor 1, and is, for example, a microcontroller. The information processing unit 4 has a communication unit 41, a storage unit 42, and a processor 43, and these components are electrically connected within the information processing unit 4 via a communication bus 40. Each component will be described in more detail.

[0025] The communication unit 41 is configured to transmit various electrical signals from the information processing unit 4 to external components. The communication unit 41 is also configured to receive various electrical signals from external components to the information processing unit 4. Specifically, the communication unit 41 receives the photocurrent J output from the AI ​​input unit 22 in the detection unit 2. The communication unit 41 outputs spatial physical quantities estimated based on the trained model 5. More preferably, the communication unit 41 has a network communication function, enabling the communication of various information between the sensor 1 and external devices via a network such as the Internet.

[0026] The memory unit 42 stores various types of information as defined above. This can be implemented, for example, as a storage device such as a solid-state drive (SSD) that stores various programs related to the sensor 1 executed by the processor 43, or as a memory such as random access memory (RAM) that stores temporarily necessary information (arguments, arrays, etc.) related to program calculations. The memory unit 42 stores various programs and variables related to the sensor 1 executed by the processor 43. In particular, in this embodiment, the memory unit 42 stores the trained model 5.

[0027] The processor 43 is, for example, a central processing unit (CPU) not shown. The processor 43 realizes various functions related to the sensor 1 by reading predetermined programs stored in the memory unit 42. That is, information processing by software stored in the memory unit 42 is concretely realized by the processor 43, which is an example of hardware, and can be executed as each functional unit included in the processor 43. These will be described in more detail in the next section. Note that the processor 43 is not limited to a single unit, and may be implemented with multiple processors 43 for each function, or a combination thereof.

[0028] In this configuration, the spatial physical quantities of the object Ob can be measured using only the sensor 1 without the need for a separate computer or the like.

[0029] 2. Functional Configuration This section describes the functional configuration of this embodiment. As mentioned above, information processing by software stored in the memory unit 42 is concretely realized by the processor 43, which is an example of hardware, and each functional unit included in the processor 43 can be executed.

[0030] Figure 3 is a block diagram showing the functions realized by the processor 43 and other components in the sensor 1. Specifically, the processor 43 comprises, as functional units, an acquisition unit 431, a conversion unit 432, an input processing unit 433, and an output unit 434.

[0031] The acquisition unit 431 is configured to acquire various information received from the outside via the communication unit 41 or previously stored in the storage unit 42 as an acquisition step. For example, the acquisition unit 431 acquires the photocurrent J output from the photodetector 32.

[0032] The conversion unit 432 is configured to perform predetermined calculations on various pieces of information acquired by the acquisition unit 431 as a conversion step to convert the information. For example, the conversion unit 432 may calculate a predetermined value based on multiple photocurrent J values ​​acquired by the acquisition unit 431 as the converted information.

[0033] The input processing unit 433 is configured to input input parameters to the trained model 5 stored in the storage unit 42 as an input processing step.

[0034] The output unit 434 is configured to output various information as an output step. Specifically, the output unit 434 may output spatial physical quantities estimated based on the trained model 5, such as distance d and angles θ and φ.

[0035] 3. Configuration of the pre-trained model 5 This section describes the trained model 5 stored in the memory unit 42. The trained model 5 is a model that has been pre-trained to learn the relationship between spatial physical quantities relating to the reference plane of the sensor 1 and the object Ob, and the value or calculated value of the photocurrent J. Preferably, the trained model 5 should be trained to learn the relationship between the value or a predetermined calculated value of the photocurrent J and the spatial physical quantities relating to the object Ob and the reference plane of the sensor 1 for each material of the object Ob. Specifically, examples include object Ob made of specularly reflective material, object Ob made of diffusely reflective white material, object Ob made of green rigid PVC sheet, object Ob made of yellow rigid PVC sheet, object Ob made of blue rigid PVC sheet, object Ob made of aluminum sheet, object Ob made of rough-finished aluminum sheet, object Ob made of wire mesh, object Ob made of leather, etc. According to this embodiment, measurements can be robustly performed even on objects Ob that are conventionally difficult to measure, such as objects Ob with a mirrored surface, transparent objects Ob, or objects Ob with irregularities on their surface.

[0036] Figure 4 is a schematic diagram showing the configuration of the neural network in the trained model 5. As shown in Figure 4, the trained model 5 comprises an input layer 51, an intermediate layer 52 consisting of at least one layer, and an output layer 53.

[0037] The input layer 51 receives, as an input signal, the value of the photocurrent J output from each photodetector 32, or a calculated value obtained by performing a predetermined operation on the photocurrent J. In Figure 4, at least eight input terminals (white circles in the figure) are shown as the input layer 51, but this is merely an example and is not limited to this.

[0038] The hidden layer 52 is the layer that performs the main processing of the neural network, and its number is not particularly limited. Multiple neurons in each layer (white circles in the diagram) are connected to multiple neurons in the next layer. For example, one neuron in the first layer of the hidden layer 52 receives weighted input from each terminal of the input layer 51, and sends the calculation result to each neuron in the second layer of the hidden layer 52, each with its own weight. Subsequently, one neuron in the second layer of the hidden layer 52 receives weighted input from each neuron in the first layer of the hidden layer 52, and sends the calculation result to each neuron in the third layer of the hidden layer 52, each with its own weight. This process is executed sequentially according to the number of layers in the hidden layer 52.

[0039] The output layer 53 is a layer that outputs the output results of the trained model 5. In Figure 4, two output terminals (white circles in the figure) are shown as the output layer 53, but this is merely an example and is not limited to this. Preferably, the distance d and angles θ and φ are output as spatial physical quantities relating the reference plane and object Ob in the sensor 1. That is, in relation to the trained model 5, the processor 43 in the information processing unit 4 is configured to input the value of the photocurrent J or a calculated value as an input parameter to the trained model 5 and to measure spatial physical quantities based on the output results from the trained model 5.

[0040] 4. Flow of measurement processing including information processing method This section describes the flow of the measurement process for object Ob using sensor 1. Figure 5 is an activity diagram showing the flow of the measurement process for object Ob using sensor 1. Figure 6 is a schematic diagram showing the timing of light emission from each light-emitting element 31 and the manner in which light is received by the light-receiving element 32, with the light-emitting elements 31 emitting light in the order of Figure 6A, Figure 6B, Figure 6C, and Figure 6D. The flow of the measurement process will be described below in accordance with this activity diagram.

[0041] In this embodiment, a sensor 1 is assumed to have four light-receiving and light-emitting blocks 3, and each light-emitting element 31 in the light-receiving and light-emitting blocks 3 emits light separately at different timings. With this configuration, the reflected light L originating from each light-emitting element 31 can be easily distinguished. Hereinafter, we will express this as "the k-th (k=1,2,3,4) light-emitting element 31 emits light." That is, light-emitting elements 31a, 31b, 31c, and 31d each emit light periodically at different timings. First, with k=1 (activity A001), the first light-emitting element 31, for example, light-emitting element 31a as shown in Figure 6A, emits light (activity A002).

[0042] Next, light emitted from the light-emitting element 31a is irradiated onto the object Ob, and the reflected light L is incident on multiple light-receiving elements 32. Specifically, light-receiving element 32a receives the reflected light L originating from the light-emitting element 31a as the primary light (Activity A003). Note that light-receiving element 32a is a pair of light-receiving elements 32 with the light-emitting element 31a, and is located in the same light-receiving block 3a as the light-emitting element 31a. In addition, light-receiving element 32b receives the reflected light L originating from the light-emitting element 31a as crosstalk light (Activity A004). Note that light-receiving element 32b is a light-receiving element 32 located in the light-receiving block 3b to the left of the light-receiving block 3a where the light-emitting element 31a is located. In addition, light-receiving element 32d receives the reflected light L originating from the light-emitting element 31a as crosstalk light (Activity A005). The light-receiving element 32d is a light-receiving element 32 located in the light-receiving block 3d, which is located to the right of the light-receiving block 3a where the light-emitting element 31a is located. Then, a photocurrent J corresponding to the received reflected light L is output from each of the light-receiving elements 32a, 32b, and 32d. Subsequently, the acquisition unit 431 in the information processing unit 4 acquires each photocurrent J. In other words, the light-receiving element 32 receives the reflected light L originating from the associated light-emitting element 31 as the primary light, and receives the other reflected light L as crosstalk light in a way that can be distinguished from the primary light.

[0043] Next, since k < 4, the value of k is incremented (activity A006), and the second light-emitting element 31, for example, light-emitting element 31b as shown in Figure 6B, emits light (activity A002).

[0044] Next, light emitted from the light-emitting element 31b is irradiated onto the object Ob, and the reflected light L is incident on multiple photodetectors 32. Specifically, photodetector 32b receives the reflected light L originating from the light-emitting element 31b as the primary light (Activity A003). Photodetector 32c receives the reflected light L originating from the light-emitting element 31b as crosstalk light (Activity A004). Photodetector 32a receives the reflected light L originating from the light-emitting element 31b as crosstalk light (Activity A005). Then, a photocurrent J corresponding to the received reflected light L is output from each of the photodetectors 32b, 32c, and 32a. Subsequently, the acquisition unit 431 in the information processing unit 4 acquires each photocurrent J.

[0045] Next, since k < 4, the value of k is incremented (activity A006), and the third light-emitting element 31, for example, light-emitting element 31c as shown in Figure 6C, emits light (activity A002).

[0046] Next, light emitted from the light-emitting element 31c is shone onto the object Ob, and the reflected light L is incident on multiple photodetectors 32. Specifically, photodetector 32c receives the reflected light L originating from the light-emitting element 31c as the primary light (Activity A003). Photodetector 32d receives the reflected light L originating from the light-emitting element 31c as crosstalk light (Activity A004). Photodetector 32b receives the reflected light L originating from the light-emitting element 31c as crosstalk light (Activity A005). Then, a photocurrent J corresponding to the received reflected light L is output from each of the photodetectors 32c, 32d, and 32b. Subsequently, the acquisition unit 431 in the information processing unit 4 acquires each photocurrent J.

[0047] Next, since k < 4, the value of k is incremented (activity A006), and the fourth light-emitting element 31, for example, the light-emitting element 31d as shown in Figure 6D, emits light (activity A002).

[0048] Next, light emitted from the light-emitting element 31d is shone onto the object Ob, and the reflected light L is incident on multiple photodetectors 32. Specifically, photodetector 32d receives the reflected light L originating from the light-emitting element 31d as the primary light (activity A003). Photodetector 32a receives the reflected light L originating from the light-emitting element 31d as crosstalk light (activity A004). Photodetector 32c receives the reflected light L originating from the light-emitting element 31d as crosstalk light (activity A005). Then, a photocurrent J corresponding to the received reflected light L is output from each of the photodetectors 32d, 32a, and 32c. Subsequently, the acquisition unit 431 in the information processing unit 4 acquires each photocurrent J. As a result of the four light emission cycles described above, 12 channels of photocurrent J have been acquired.

[0049] In this case, since k=4, the processing of activities A007 to A009 is executed. Specifically, first, the conversion unit 432 performs preprocessing to convert each photocurrent J acquired by the acquisition unit 431 in the information processing unit 4 into a predetermined calculated value (activity A007). The predetermined calculated value is not particularly limited as long as it is a value calculated based on each photocurrent J. The number of channels in such a calculated value may be the same as the number of channels of the photocurrent J (12 in one example), or it may be less.

[0050] Next, the input processing unit 433 in the information processing unit 4 inputs each calculated value obtained in activity A007 as input parameters to the trained model 5 (activity A008).

[0051] Finally, the output unit 434 in the information processing unit 4 outputs the spatial physical quantities estimated from the trained model 5, in this case the distance d and angles θ and φ (activity A009). Since the above processing is performed continuously as long as the sensor 1 is operating, the distance d and angles θ and φ are measured sequentially according to the operating frequency of the sensor 1. The time from the emission of light from the first light-emitting element 31 until the distance d and angles θ and φ are estimated to be measured is often 100 milliseconds or less, preferably 10 milliseconds or less, more preferably 1 millisecond or less, and even more preferably 0.7 milliseconds or less. The time until measurement is, specifically, for example, 100, 95, 90, 85, 80, 75, 70, 65, 60, 55, 50, 45, 40, 35, 30, 25, 20, 15, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3, 0.2, 0.1 milliseconds, and may be within the range of any two of the values ​​exemplified here. That is, the time from the emission of light from the k-th light-emitting element 31 to the emission of light from the (k+1)-th light-emitting element 31 should be less than or equal to 1 / 4 of the above measurement time. More generally, the time from the emission of light from the k-th light-emitting element 31 to the emission of light from the (k+1)-th light-emitting element 31 should be less than or equal to the time obtained by dividing "the time from the emission of light from the first light-emitting element 31 until the distance d and angles θ, φ can be estimated and measured" by "the number of light-emitting elements 31".

[0052] In summary, the sensor 1 according to this embodiment is a sensor for measuring an object Ob. The sensor 1 comprises a plurality of light-emitting elements 31, a light-receiving element 32, and an AI input unit 22. The light-emitting elements 31 are each provided at different positions on the substrate 21. The light-receiving element 32 is provided on the substrate 21. The light-receiving element 32 is configured to receive one of the reflected light L originating from each light-emitting element 31 as the main light, and to receive the other reflected light L as crosstalk light in a way that is distinguishable from the main light. The reflected light L is light emitted from each light-emitting element 31 and reflected from the object Ob, respectively. Based on the main light and crosstalk light that are received in a way that is distinguishable, the sensor 1 measures spatial physical quantities relating to the reference plane and the object Ob in the sensor 1.

[0053] From another perspective, this information processing method comprises the following steps: In the acquisition step, the photocurrent J (an example of an electrical physical quantity) output from the photodetector 32 is acquired. In the input processing step, the value of the photocurrent J or a calculated value based on the photocurrent J is input to the trained model 5 as an input parameter. The trained model 5 is a model that has been pre-trained to learn the relationship between spatial physical quantities relating to the reference plane and the object Ob in the sensor 1 and the value of the photocurrent J or a calculated value. In the output step, the spatial physical quantities are estimated and output based on the trained model 5.

[0054] This configuration enables the realization of a sensor with a simpler structure and higher accuracy. In particular, instead of using only the reflected light L between corresponding elements, such as the light-emitting element 31a and the light-receiving element 32a, for measurement, crosstalk light, which was conventionally considered a disturbance, is also used as an input parameter for measurement judgment. This allows for obtaining more diverse information and improving measurement accuracy.

[0055] [others] The sensor 1 according to this embodiment may be further modified or improved.

[0056] The light-emitting element 31 and the light-receiving element 32 do not have to be arranged in a one-to-one ratio. Figure 7 is a schematic front view showing the configuration of a modified sensor 1. The modified sensor 1 has, as a minimum configuration, two light-emitting elements 31x and 31y and one light-receiving element 32x. Light emitted from the light-emitting element 31x is reflected off the object Ob, and the reflected light L1 (an example of the main light) is received by the light-receiving element 32x. Also, light emitted from the light-emitting element 31y is reflected off the object Ob, and the reflected light L2 (an example of crosstalk light) is received by the light-receiving element 32x. Even with a sensor 1 of this configuration, the distance d and angle θ (angle in one direction) can be estimated using a pre-trained model 5 that has been machine-learned.

[0057] In the above-described embodiment, the example was given in which the light-emitting / receiving block 3 includes a light-emitting element 31 and a light-receiving element 32 in a one-to-one ratio, and each of the light-emitting / receiving blocks 3 is arranged on the substrate 21 in an annular and equally spaced manner. However, the light-emitting element 31 and the light-receiving element 32 may be arranged on the substrate 21 without any regularity whatsoever. Even in such a case, the distance d and angles θ and φ can be estimated by preparing a pre-trained model 5 that has been subjected to machine learning.

[0058] Although the sensor 1 has been described as including a detection unit 2 and an information processing unit 4, the sensor 1 may also consist only of a detection unit 2, and the measurement results may be output using a computer (not shown) connected to it. In such a case, a learned model 5 may be stored in a memory unit (not shown) built into the computer (not shown), and input processing to the learned model 5 may be performed by a processor (not shown). Furthermore, the mode of connection is not particularly limited; they may be connected directly by wires, or the sensor 1 may have a wireless communication function, and the sensor 1 and the computer (not shown) may be connected via wireless communication. In such a case, one computer (not shown) may be used to support multiple sensors 1.

[0059] The aforementioned sensor 1 may be manufactured by a predetermined manufacturing method. Such a manufacturing method may include the step of arranging a plurality of light-emitting elements 31 and at least one light-receiving element 32 on a substrate 21. For example, the light-emitting elements 31 may be arranged on the substrate 21, followed by the arrangement of the light-receiving element 32, and then the output of the light-receiving element 32 may be connected to an information processing unit 4 such as a microcontroller. By such a method, a sensor with a simpler structure and high accuracy can be manufactured.

[0060] In the above-described embodiment, the resulting photocurrents J were distinguished by each light-emitting element 31 emitting light at different timings. However, instead of timing, the resulting photocurrents J may be distinguished by blinking each light-emitting element 31 at different frequencies or by using different wavelengths of light emitted by the light-emitting elements 31.

[0061] The system may be configured so that not only the left and right adjacent light-receiving elements 32, but also opposing light-receiving elements 32 receive crosstalk light. In other words, the position of the light-receiving elements 32 that receive crosstalk light can be freely determined according to the number of light-receiving and light-emitting blocks 3.

[0062] The trained model 5 may be a one-dimensional convolutional network or a two-dimensional convolutional network instead of a neural network.

[0063] The electrical and physical quantities generated from the light-receiving element 32 are not limited to the photocurrent J, but may also be voltage, for example.

[0064] Furthermore, they may be provided in the following embodiments.

[0065] (1) A sensor for measuring an object, comprising a plurality of light-emitting elements and at least one light-receiving element, wherein the light-emitting elements are each provided at different positions on a substrate, and the light-receiving element is provided on the substrate and receives one of the reflected light originating from each light-emitting element as the primary light, and receives the other reflected light as crosstalk light in a manner distinguishable from the primary light, wherein the reflected light is light emitted from each of the light-emitting elements and reflected from the object, respectively, and measures a spatial physical quantity relating to the reference plane of the sensor and the object based on the primary light and the crosstalk light that are received in a manner distinguishable from the primary light.

[0066] This configuration enables the realization of a sensor with a simpler structure and higher accuracy.

[0067] (2) The sensor described in (1) above, wherein the spatial physical quantity includes at least the distance between the reference plane and the object and the angle of the object with respect to the reference plane.

[0068] This configuration allows for the measurement of the distance and angle to an object, and can be applied to various situations such as object picking and inspection.

[0069] (3) The sensor described in (2) above, wherein the distance is 50 mm or less.

[0070] This configuration allows for accurate identification of objects located at very close range.

[0071] (4) A sensor described in any one of (1) to (3) above, wherein the number of light-receiving elements is the same as the number of light-emitting elements, and each light-receiving element is associated with the light-emitting element in a one-to-one correspondence, and the reflected light originating from the associated light-emitting element is received as the main light, and the reflected light other than the main light is received as crosstalk light in a manner that can be distinguished from the main light.

[0072] In this configuration, light can be obtained from multiple light-receiving elements, thereby improving the accuracy of the sensor.

[0073] (5) In the sensor described in (4) above, the light receiving element and the corresponding light-emitting element are adjacent to each other to form a light-receiving block, and a plurality of such light-receiving blocks are provided on the substrate.

[0074] In this configuration, light can be obtained from multiple light-receiving elements, thereby improving the accuracy of the sensor.

[0075] (6) In the sensor described in (5) above, each of the light-receiving and light-emitting blocks is arranged on the substrate in an annular shape and at equal intervals.

[0076] In this configuration, symmetry arises in each physical quantity corresponding to the light obtained from multiple photodetectors, thereby improving the accuracy of the sensor.

[0077] (7) A sensor described in any one of (1) to (6) above, wherein the number of light-emitting elements is 2 to 4.

[0078] According to this embodiment, a sensor with an even simpler structure and high accuracy can be realized.

[0079] (8) A sensor described in any one of (1) to (7) above, wherein each of the light-emitting elements emits light separately at different timings.

[0080] According to this embodiment, reflected light originating from each light-emitting element can be easily distinguished.

[0081] (9) A sensor described in any one of (1) to (8) above, wherein the light receiving element is configured to generate an electrophysical quantity corresponding to the reflected light, and measures a spatial physical quantity relating to the reference plane and the object in the sensor based on the electrophysical quantities based on the main light and the crosstalk light, respectively, which are received in a distinguishable manner.

[0082] This configuration enables electrical handling essential for information processing, resulting in the realization of high-precision sensors.

[0083] (10) The sensor described in (9) above, further comprising an AI input unit, wherein the AI ​​input unit inputs the value of the electrical physical quantity or a calculated value based on the electrical physical quantity as an input parameter to a trained model, wherein the trained model is a model that has been pre-trained to understand the relationship between the spatial physical quantity relating to the reference plane and the object in the sensor and the value of the electrical physical quantity or the calculated value.

[0084] According to this configuration, a highly accurate sensor based on machine learning can be realized.

[0085] (11) The sensor described in (10) above further comprises a storage unit and a processor, wherein the storage unit stores the trained model, and the processor is configured to input the value of the electrical physical quantity or the calculated value as an input parameter to the trained model, and to measure the spatial physical quantity based on the output result from the trained model.

[0086] In this configuration, the spatial physical quantities of an object can be measured using only the sensor without the need for a separate computer or other device.

[0087] (12) A sensor described in any one of (1) to (11) above, wherein the light-receiving element is a light-receiving element having broad directivity.

[0088] This configuration allows for reliable reception of both the primary light and crosstalk light, thereby improving the accuracy of the sensor.

[0089] (13) A method for manufacturing a sensor, comprising the step of arranging a plurality of light-emitting elements and at least one light-receiving element on a substrate, wherein the sensor is the sensor described in any one of (1) to (12) above.

[0090] This method allows for the manufacture of sensors with a simpler structure and higher precision.

[0091] (14) An information processing method comprising the following steps: an acquisition step in which a photocurrent output from a photodetector is acquired; an input processing step in which the value of the electrical physical quantity or a calculated value based on the electrical physical quantity is input to a trained model as an input parameter, wherein the trained model is a model that has been pre-trained to learn the relationship between a spatial physical quantity relating to a reference plane and an object in a sensor and the value of the electrical physical quantity or the calculated value; and an output step in which the spatial physical quantity is estimated and output based on the trained model.

[0092] This configuration enables the realization of a sensor with a simpler structure and higher accuracy. Of course, this is not always the case.

[0093] Finally, various embodiments of the present invention have been described, but these are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of symbols]

[0094] 1: Sensor 2: Detection unit 21: Base material 22:AI input section 3: Light-receiving block 3a: Light-receiving block 3b: Light-receiving block 3c: Light-receiving block 3D: Light-receiving block 31: Light-emitting element 31a: Light-emitting element 31b: Light-emitting element 31c: Light-emitting element 31d: Light-emitting element 31x: Light-emitting element 31y: Light-emitting element 32: Photodetector 32a: Photodetector 32b: Photodetector 32c: Photodetector 32d: Photodetector 32x: Photodetector 4: Information Processing Department 40: Communications bus 41: Communications Department 42: Storage section 43: Processor 431: Acquisition Department 432: Conversion section 433: Input Processing Unit 434: Output section 5: Pre-trained models 51: Input Layer 52: Middle Class 53: Output Layer J: Photocurrent L: Reflected light L1: Reflected light L2: Reflected light Ob: object d: distance θ: angle φ: Angle

Claims

1. A sensor for measuring an object, It comprises multiple light-emitting elements and at least one light-receiving element, The light-emitting elements are provided at different positions on the substrate, The aforementioned light-receiving element is Provided on the aforementioned substrate, One of the reflected light sources originating from each light-emitting element is received as the primary light, and the other reflected light is received as crosstalk light. Based on the received primary light, the crosstalk light, and a predetermined learned model, the primary light and the crosstalk light are distinguished, and spatial physical quantities relating to the reference plane and the object in the sensor are measured, and here, The reflected light is the light emitted from each of the light-emitting elements and reflected from the object, respectively. The aforementioned trained model is a model that has been pre-trained to learn the relationship between the spatial physical quantity and the value determined by the primary light and the crosstalk light.

2. In the recovery described in claim 1, The spatial physical quantity includes at least the distance between the reference plane and the object, and the angle of the object with respect to the reference plane.

3. In the detection described in claim 2, The aforementioned distance is 50 mm or less.

4. In the recovery described in claim 1, The aforementioned light-receiving element is The number of these light-emitting elements is the same as the number of the aforementioned light-emitting elements, and each of the aforementioned light-emitting elements is associated with a one-to-one correspondence. A device configured to receive the reflected light originating from the associated light-emitting element as the primary light, and to receive the reflected light other than the primary light as crosstalk light.

5. In the sensor according to claim 4, The light-receiving element and the corresponding light-emitting element form a light-receiving block adjacent to each other. A plurality of the aforementioned light-receiving and light-emitting blocks are provided on the substrate.

6. In the detection described in claim 5, Each of the light-receiving and light-emitting blocks is arranged on the substrate in an annular shape and at equal intervals.

7. In the recovery described in claim 1, The number of light-emitting elements is 2 to 4.

8. In the recovery described in claim 1, Each of the aforementioned light-emitting elements emits light separately at different timings.

9. In the recovery described in claim 1, The light-receiving element is configured to generate an electrophysical quantity corresponding to the reflected light, This device measures spatial physical quantities relating to the reference plane and the object in the sensor, based on the received main light and the electrophysical quantities derived from the crosstalk light, respectively.

10. In the recovery described in claim 9, It also features an AI input unit, The AI ​​input unit inputs the value of the electrical physical quantity or a calculated value based on the electrical physical quantity as an input parameter to the trained model, wherein the trained model is a model that has been pre-trained to understand the relationship between the spatial physical quantity relating to the reference plane of the sensor and the object, and the value of the electrical physical quantity or the calculated value.

11. In the detection described in claim 10, It further includes a memory unit and a processor, The memory unit stores the trained model, The processor is configured to input the value of the electrical physical quantity or the calculated value as an input parameter to the trained model, and to measure the spatial physical quantity based on the output result from the trained model.

12. In the recovery described in claim 1, The aforementioned light-receiving element is a light-receiving element having broad directivity.

13. A method for manufacturing a sensor, A method comprising the step of arranging a plurality of light-emitting elements and at least one light-receiving element on a substrate, wherein the sensor is the sensor according to any one of claims 1 to 12.

Images