Apparatus and method for detecting and identifying biological or non-biological entities

Abstract

A device for detecting and identifying biological or non-biological entities, comprising: at least one detectable real entity unit (12, 22, 32, 42) configured on a real object (13, 14, 23, 24, 36, 37, 48) to be detected; and at least one detection module (140) for detecting the real entity unit, configured on another component (49). The real entity unit (12, 22, 32, 42) cyclically broadcasts a real image, a virtual image or a virtual representation of the real object to be detected and information or attributes associated with the identity of the entity in a one-way, non-interactive manner; the real entity unit (12, 22, 32, 42) has a unique digital identity, including a storage and / or digital recording device, an internal power supply device, a remote receiving device, a collection device and a broadcasting device; the detection module includes a receiving device, a temporary storage device, a display and / or listening device and a processing device.

Description

Technical Field

[0001] This invention relates to the field of physical detection technology, and more specifically, to an apparatus and method for detecting and identifying biological or non-biological entities. Background Technology

[0002] Currently, entity detection and recognition technologies mainly include two modes. The first is direct detection and recognition through human vision, which relies on direct observation by the human eye and identification and judgment by the human brain. The second is detection and recognition using technical means associated with information processing devices, which utilizes various sensors connected to computers, such as radar, lidar, cameras, ultrasonic sensors, and RFID radio tag chips, combined with data processing algorithms to achieve entity detection and recognition.

[0003] However, all of the above detection modes have certain limitations. Regarding direct detection and recognition by human vision, when objects are far away and / or in adverse weather conditions such as fog, rain, or nighttime, it is difficult for the naked eye to effectively detect, observe, identify, and recognize these objects.

[0004] For detection and identification using computer-related technologies, various sensors also have inherent limitations. Radar cannot distinguish the color and outline of objects; ultrasonic sensors have a very short operating range, typically less than ten meters; lidar is not sensitive to color and is easily affected by adverse weather conditions such as rain, snow, and dense fog; RFID chips also have a limited operating range, typically less than two hundred meters; and cameras are highly sensitive to adverse environments such as rain, snow, dense fog, and lighting conditions. Furthermore, the identification algorithms associated with information processing systems are not entirely reliable; their training requires large amounts of data and relies on powerful computers.

[0005] Besides the limitations of the detection methods themselves, the objects to be identified also have many shortcomings. For example, objects carrying advertising information, such as billboards, displays, advertising notices, and signs, can cause visual pollution in cities, towns, and roads. This leads to prohibitions or restrictions on their installation at town and village entrances or inside areas, or limitations on their size, quantity, and installation on streets and buildings. Furthermore, passengers at the rear of wheeled vehicles cannot see these billboards and signs. Summary of the Invention

[0006] In light of the above, the applicant conducted research aimed at facilitating the detection and identification of biological or non-biological entities in order to address the shortcomings of existing technologies.

[0007] These studies have led to the design of devices and methods that ensure the transformation of biological or non-biological entities into detectable objects to facilitate their detection, identification, and recognition, thereby optimizing the use of their identification information for display on fixed or mobile digital screens, or for the guidance and control of semi-autonomous wheeled or flying machines or autonomous wheeled or flying machines, or for the use of their identification information to understand the possible modes of motion of biological or non-biological entities when they are in motion.

[0008] According to the present invention, an apparatus for detecting and identifying biological or non-biological entities includes at least one detectable physical entity unit and at least one detection module.

[0009] The real entity unit is integrated, embedded, or bound to the real object to be detected, and can also partially or entirely replace the real object to be detected.

[0010] The real entity unit sends out real images, virtual images, or virtual representations of the real object to be detected, as well as information or attributes related to the identity of the real object to be detected, in a one-way, non-interactive, and cyclical broadcast manner.

[0011] The real entity unit has a unique digital identity and includes:

[0012] A storage and / or digital recording device for storing a real or virtual image of a real object to be detected and / or information or attributes related to the identity of the real object to be detected;

[0013] An internal power supply unit is used to power the actual physical unit.

[0014] A remote receiving device is used to remotely receive information or attributes associated with a real object to be detected, and updated data of a real or virtual image of the real object to be detected.

[0015] The acquisition device is used to acquire real images of the real object to be detected, and the real images may or may not include images of the surrounding environment of the real object to be detected.

[0016] A broadcasting device for cyclically broadcasting data to one or more mobile or fixed detection modules;

[0017] Each of the detection modules includes:

[0018] A receiving device is configured to receive image and / or information data broadcast by the real entity unit;

[0019] A temporary storage device for temporarily storing the data;

[0020] Display and / or listening device for displaying and / or listening to the data;

[0021] A processing device used to process data.

[0022] The following list provides examples of real-world objects to be detected in a non-exhaustive manner, with some instances accompanied by relevant information:

[0023] Biological entities:

[0024] Humans:

[0025] Male + Child / Adult / Elderly; Female + Child / Adult / Elderly.

[0026] Animals: dog + breed; cat; cow; sheep; etc.

[0027] Non-living entities:

[0028] bike;

[0029] Motorcycles + Motorcycle Brands;

[0030] Car + Car Brand + Size;

[0031] Other wheeled machines + dimensions;

[0032] Aircraft + Aircraft type + Size;

[0033] Train + Train Type + Size;

[0034] Other flying machines + dimensions;

[0035] building;

[0036] shop;

[0037] neighborhood;

[0038] sidewalk;

[0039] Road traffic signs + sign type;

[0040] Advertising signage + its content;

[0041] Billboard + its content;

[0042] Advertising preview sign + its content;

[0043] pedestrian crossing.

[0044] When the real entity unit performs cyclic transmission, it transmits detectable data in addition to visual data, thereby enabling the detection and identification of relevant real entities. Detection and identification can be completed without visual observation of the entity. Furthermore, since the real entity unit possesses signal transmission behavior, highly precise detection methods are not required.

[0045] Therefore, the device of the present invention overcomes the shortcomings of the prior art.

[0046] The information or attributes broadcast by the device that are related to the identity of the real object to be detected can be of various types. The information or attributes related to the identity of the real object to be detected include: object type, object properties, object name, object content, object size, and object motion probability.

[0047] "Broadcast" is a well-known term in the field, referring to one-way broadcasting to multiple receivers; in this solution, the receiver is the detection module, and the number of receivers is unlimited.

[0048] Furthermore, the broadcast is a television broadcast.

[0049] Furthermore, the broadcast is a radio broadcast used to transmit audio explanations of the type of real object to be detected, as well as information or attributes related to the identity of the real entity to be detected.

[0050] Audio narration constitutes part of the information or attributes associated with an entity.

[0051] Radio broadcasting requires less energy and has a narrower broadcast bandwidth than television broadcasting.

[0052] Furthermore, the information or attributes broadcast that are related to the identity of the real entity to be detected include the entity's location, object posture, direction of movement, and speed of movement. This type of information can be broadcast in audio form.

[0053] Furthermore, the storage capacity of the storage device and / or digital recording device is 0.1 megabytes to 10 terabytes or more.

[0054] Furthermore, the power supply device is rechargeable.

[0055] Furthermore, the power supply device is connected to an external charging device, which includes a solar panel, a wind turbine, a magnetic field power generation device, a piezoelectric power generation device, and a power grid.

[0056] Furthermore, the communication distance of the remote receiving device is 500 meters to 100,000 kilometers, and the remote receiving device is implemented using a digital wireless link or a wired link.

[0057] Furthermore, the device includes a conversion device for converting a real physical image into a virtual image to protect the privacy of the real physical image.

[0058] Furthermore, the execution cycle of the television broadcasting loop and / or the radio broadcasting loop is 1 to 120 seconds, and the transmission distance can reach an ultra-long distance of more than 500 meters to 20,000 meters; the data transmission rate of the television broadcasting and / or radio broadcasting can be between 100 kilobytes / second and 10 gigabytes / second and above.

[0059] Furthermore, the detection module is embedded or integrated in wheeled and flying machines, autonomous or semi-autonomous wheeled and flying machines, desktop computers, laptops, smartphones, or carried by the human body.

[0060] Furthermore, the real entity unit is equipped with a detection device, which is used to detect the object's position, orientation, direction of motion, and speed. The detection device consists of a GPS sensor, an inertial gyroscope sensor, and an accelerometer.

[0061] Therefore, billboard-like entities can be divided into two parts: the technical carrier of advertising information and the advertising information itself. This invention simplifies the technical carrier of advertising information to a simple pillar, or even uses other entities such as trees as substitutes or substitutes for the technical carrier. The trees serve as supporting components for the physical unit, which then broadcasts the advertising information. This solution can eliminate visual pollution associated with installing billboards, advertising announcement signs, or signage at the entrances or inside towns and villages, or along land roads and railway lines.

[0062] The same applies to road traffic signs.

[0063] Considering that it is intended for human use, the detection module and / or physical unit can be integrated into backpacks, handbags, suitcases, bicycle helmets, and motorcycle helmets.

[0064] The detection module and / or physical unit can be based on a mobile terminal such as a smartphone; after modifying the smartphone, the new functions of this device can be realized. For example, the relevant functions can be controlled through a dedicated application, and the wheeled vehicle driver program and vehicle safety program are compatible with the mobile terminal.

[0065] In the guidance, autonomous guidance and control scenarios of autonomous or semi-autonomous equipment, the information received by the detection module from the real entity unit can be used to manipulate the actuators of the equipment to perform actions such as steering, acceleration or braking.

[0066] Furthermore, the detection module also includes a timing device, which is used to determine the connection duration between the detection module and the real physical unit.

[0067] Furthermore, the data processing device used to process the data from the detection module can communicate and interact with the guidance and control systems of flying and wheeled machines.

[0068] Another object of the present invention is to provide a method of using the above-described device, comprising:

[0069] A storage and / or digital recording device is used to store real images, virtual images, and information or attributes related to the identity of the real objects to be detected for various types of real objects, and / or to store and save audio narration voice of the type of real object to be detected and information or attributes related to its identity. The storage capacity of the storage device and / or digital recording device is in the range of 0.1 megabytes to 10 terabytes and above.

[0070] An internal power supply device is used to power the real physical unit. The internal power supply device may be rechargeable or non-rechargeable and may be connected to an external charging device, which may include a solar panel, a wind turbine, a magnetic field device, a piezoelectric device, or a power grid.

[0071] A remote receiving device is used to remotely receive information or attributes related to the identity of a real object to be detected, or updated data of a real or virtual image, via a digital wireless link or wired link, with a receiving range of 500 meters to 100,000 kilometers.

[0072] Acquisition and conversion devices are used to acquire real images of real objects to be detected, with or without relevant environmental information, and convert the real images into virtual images to ensure the confidentiality of the real images.

[0073] The detection device is used to acquire the position, posture, direction of motion, and speed of the real object to be detected.

[0074] A broadcasting device for transmitting, without bidirectional interactive communication, and in a cyclical manner, a television broadcast of a real image, virtual image, or virtual representation of a real object to be detected, as well as information or attributes related to the identity of the real object to be detected, with or without the object's position, posture, direction of motion, and speed; and / or a radio broadcast of the audio of the real object to be detected, as well as information or attributes related to its identity, with or without the object's position, posture, direction of motion, and speed.

[0075] The execution cycle of the television broadcasting and / or radio broadcasting loop is from 1 second to 120 seconds, and the transmission distance is from 500 meters to 20,000 meters;

[0076] The system broadcasts a real image, virtual image, or virtual representation of the real object to be detected, along with identity-related information or attributes, to one or more detection modules via television in a one-way transmission, without bidirectional interactive communication and in a loop. The television broadcast content may or may not include the object's position, posture, direction of motion, and speed. Alternatively, the system broadcasts an audio narration of the real object to be detected, along with identity-related information or attributes, to one or more detection modules via radio in a one-way transmission, without bidirectional interactive communication and in a loop. The radio broadcast content may or may not include the object's position, posture, direction of motion, and speed.

[0077] The detection module is embedded or integrated into portable terminals such as wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines, autonomous / semi-autonomous flying machines, desktop computers, and smartphones, or carried by the human body. The detection module includes...

[0078] A receiving device is used to receive a real or virtual image of a real object to be detected, which includes identity-related information or attributes, and / or to receive audio narration of the real object to be detected and information or attributes related to the identity of the real object to be detected.

[0079] A storage device for temporarily storing a real or virtual image of a real object to be detected, along with identity-related information or attributes, and / or audio narration of the real object to be detected, as well as information or attributes associated with the identity of the real object to be detected.

[0080] Display or projection device for displaying or projecting a real or virtual image of a real object to be detected, which is associated with information or attributes related to its identity, and / or listening device for listening to an audio narration of the real object to be detected and information or attributes related to the identity of the real object to be detected.

[0081] The processing device is used to perform calculations on real images and virtual images of real objects to be detected, which are accompanied by identity-related information or attributes.

[0082] The detection module connects to one or more real entity units simultaneously, with a single connection duration ranging from 2 seconds to 10 minutes. The detection module receives and displays real or virtual images of the real object to be detected, along with identity-related information or attributes, on its built-in display device. The information received and displayed by the detection module may or may not include the object's position, posture, direction of motion, and speed. The detection module also sends the identified type of real object and identity-related information or attributes to the control and guidance system of a semi-autonomous or autonomous wheeled / flying machine. The information sent may or may not include the object's position, posture, direction of motion, and speed.

[0083] and / or the detection module is connected to one or more real entity units at the same time, with a single connection duration ranging from 2 seconds to 10 minutes; the detection module receives the audio commentary speech and identity-related information or attributes of the real object to be detected, and outputs the audio commentary speech to the listening device matched with the detection module; and / or transmits the type and identity-related information or attributes of the identified real object to the audio information processing system, so as to realize the control and guidance of semi-autonomous / autonomous wheeled machines and flying machines; the audio commentary speech and identity-related information or attributes may or may not be accompanied by information on physical object position, object pose, motion direction and motion speed.

[0084] For the scenario where billboards, advertisement preview signs and identification signs are adapted by adopting an overall replacement method to modify the entity, this method uses other physical objects to replace the original real entity, such as natural trees, artificial trees, natural flowers or artificial flowers, and integrates the real entity unit onto this brand new replacement carrier.

[0085] Further, the real entity unit is configured with a detection device, which is used for detecting the object position, object pose, motion direction and motion speed.

[0086] Further, the processing device of the detection module communicates and interacts with the guidance and control system, and / or the detection module is equipped with an artificial intelligence-loaded speech recognition device and an audio processing device for processing received audio information, and the speech recognition and audio processing device can establish communication with the guidance and control systems of flying machines and wheeled machines.

[0087] Further, the detection module is equipped with a timing device, which is used for measuring and calculating the connection duration between the detection module and the real entity unit.

[0088] Further, in the method, a number of detection modules are arranged at intervals along a section of road; when a vehicle carrying a real entity unit travels on this road section, the traveling speed of the vehicle is calculated by measuring and calculating the travel duration of the vehicle passing through two detection modules.

[0089] This method can uniformly integrate and efficiently monitor the motion states of various types of objects, including living bodies, inanimate objects, movable objects and fixed objects, covering objects distributed or moving along land routes, sea routes, railways, air lanes, beside waterways and under waterways, as well as objects traveling and moving in the above-mentioned lanes and waterways.

[0090] Further, the method also includes the following steps:

[0091] Identifying all possible motion poses of the real object to be detected,

[0092] For each possible motion posture of the real object to be detected, a corresponding audio description, visual description, and / or encoding is matched.

[0093] One or more sensors are configured for each possible motion posture of the real object to be detected. These sensors are used to detect or acquire the motion state of the entity.

[0094] A real entity unit is configured for the real object to be detected. The real entity unit has built-in sensors corresponding to various motion actions that can be performed by all the identified entities.

[0095] Using the broadcasting device built into the real entity unit, descriptive information and / or encoding matching the motion of the real object to be detected collected by the sensor are broadcast to the detection module. The broadcasting device is configured to transmit the encoding or descriptive information corresponding to the motion of the real object to be detected in a one-way, non-interactive, looping television broadcasting mode and / or radio broadcasting mode. The execution cycle of the television broadcasting loop and / or radio broadcasting loop is 1 second to 120 seconds, and the transmission distance can reach an ultra-long distance of more than 500 meters to 20,000 meters. The data transmission rate of the television broadcasting and / or radio broadcasting is between 100 kilobytes / second and 10 gigabytes / second and above. The detection module is equipped with a receiving device for receiving audio or visual descriptions and / or codes associated with the entity's movement. The audio or visual descriptions are played on a sound playback device or displayed on a display device provided with the detection module, or the codes are transmitted to a processing device. The processing device retrieves a database containing codes corresponding to all movable actions, decodes the codes, and matches the corresponding actions. The audio or visual descriptions and / or codes associated with the entity's movement can be used to implement manual or autonomous automatic control of the entity's movement control mechanism.

[0096] Possible movement types include: forward, backward, left turn, right turn, driving direction, driving lane, braking, deceleration, changing lanes out, changing lanes in, driving in the opposite direction, and parking.

[0097] The applications under consideration are numerous, including: advertising placement in wheeled and / or flying machines, pedestrian-oriented advertising displays, manual guidance, autonomous guidance and motion control of wheeled and flying machines, road traffic signs, road navigation, and traffic violation monitoring and enforcement.

[0098] Furthermore, the detection module is in the same environment as the real object to be detected, or moves and operates within the same environment.

[0099] Furthermore, the environment in which the real entity to be detected and the detection module are located or active includes aircraft, land-based wheeled vehicles, sea vessels, and buildings.

[0100] Furthermore, the broadcast transmission distance for real images, virtual images, or virtual representations of the real object to be detected, as well as information or attributes related to the identity of the real object to be detected, is set according to the spatial boundary range of the environment.

[0101] Furthermore, for real images, virtual images, or virtual representations of the real object to be detected, as well as information or attributes related to the identity of the real object to be detected, the broadcast transmission distance is from 1 meter to 500 meters.

[0102] The present invention also provides an artificial intelligence automatic training device, which utilizes the above-mentioned device for detecting and identifying biological or non-biological entities to achieve automatic artificial intelligence training through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning.

[0103] Among them, wheeled vehicles, flying machines, autonomous / semi-autonomous wheeled and flying machines, portable terminals such as personal computers or smartphones, and personnel equipped with detection modules are all equipped with the following components:

[0104] Airborne artificial intelligence system, or establishing a communication connection with a remote artificial intelligence system;

[0105] One or more cameras, lidar, or any computer vision device capable of acquiring images and contours of a real object to be detected;

[0106] One or more detection modules, wherein the detection modules are connected to an airborne artificial intelligence system mounted on an autonomous or semi-autonomous wheeled or flying machine, a computer or smartphone-like portable terminal, or a human carrier, or are connected to a remote artificial intelligence system;

[0107] The autonomous and semi-autonomous wheeled and flying machines, portable terminals such as computers or smartphones, and human carriers are all connected to the recorded artificial intelligence system or remote artificial intelligence system.

[0108] The present invention also provides a method for automatic training of artificial intelligence using the above-described detection device through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning, comprising the following steps:

[0109] Step 1: Use cameras and / or lidar and / or other computer vision devices mounted on wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines and flying machines, desktop computers, laptops, smartphones or personnel carriers to acquire images of the real object to be detected, or acquire images of the real object to be detected including the surrounding environment.

[0110] Step 2: Transmit the images acquired in Step 1 to the artificial intelligence system to complete the identification and recognition of the real objects to be detected;

[0111] Step 3: The detection module receives a real image, virtual image, or virtual representation of the real object to be detected from the real entity unit, as well as information or attributes related to the identity of the real object to be detected.

[0112] Step 4: The detection module transmits the real image, virtual image, or virtual representation of the real object to be detected and / or identity-related information or attributes received in Step 3 to the artificial intelligence system;

[0113] Step 5: Under real-world conditions, verify, correct, or optimize the real object recognition results output by the AI ​​system in Step 2 in real time, and / or automatically train the AI ​​through one or more of the following methods: supervised learning, unsupervised learning, reinforcement learning, and deep learning.

[0114] The present invention also provides a method for enhancing and / or optimizing the performance of artificial intelligence using the above-described detection device, comprising the following steps:

[0115] Step 1: Use cameras and / or lidar and / or other computer vision acquisition devices mounted on wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines and flying machines, desktop computers, laptops, smartphones or personnel carriers to acquire images of the real object to be detected, or acquire images of the real object to be detected including the surrounding environment.

[0116] Step 2: Transmit the images acquired in Step 1 to the artificial intelligence system to complete the identification and recognition of the real objects to be detected;

[0117] Step 3: The detection module receives a real image, virtual image, or virtual representation of the real object to be detected from the real entity unit, as well as information or attributes related to the identity of the real object to be detected.

[0118] Step 4: The detection module transmits the real image, virtual image, or virtual representation of the real object to be detected and / or identity-related information or attributes received in Step 3 to the artificial intelligence system;

[0119] Step 5: Verify, correct, or optimize the recognition results of the real objects to be detected output by the artificial intelligence system in Step 2.

[0120] Furthermore, steps 1 through 5 can be performed in chronological order or not.

[0121] Furthermore, step 3 is performed simultaneously with step 1 and / or step 2.

[0122] Furthermore, the real image, virtual image, or virtual representation of the real object to be detected, as well as the information and attribute parameters corresponding to the identity of the real object to be detected, are transmitted via 4G broadcast, 5G broadcast, or 6G broadcast.

[0123] The present invention also provides a method for using the above-described detection device to perform artificial intelligence under real working conditions through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning, comprising the following steps:

[0124] Step 1: Use cameras and / or lidar and / or other computer vision devices mounted on wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines and flying machines, desktop computers, laptops, smartphones or personnel carriers to acquire images of the real object to be detected, or acquire images of the real object to be detected including the surrounding environment.

[0125] Step 2: Transmit the images acquired in Step 1 to the artificial intelligence system to complete the identification and recognition of the real objects to be detected;

[0126] Step 3: The detection module receives a real image, virtual image, or virtual representation of the real object to be detected from the real entity unit, as well as information or attributes related to the identity of the real object to be detected.

[0127] Step 4: The detection module transmits the real image, virtual image, or virtual representation of the real object to be detected and / or identity-related information or attributes received in Step 3 to the artificial intelligence system;

[0128] Step 5: Under real-world conditions, verify, correct, or optimize the real object recognition results output by the AI ​​system in Step 2 in real time, and / or automatically train the AI ​​through one or more of the following methods: supervised learning, unsupervised learning, reinforcement learning, and deep learning.

[0129] The present invention also provides a method for controlling and / or supervising and / or correcting and / or enhancing the operation of an artificial intelligence system and / or improving the operation of an artificial intelligence system under real-time and real-world conditions using the automatic training device described in claim 1, characterized by comprising the following steps:

[0130] Step 1: Use cameras and / or lidar and / or other computer vision devices mounted on wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines and flying machines, desktop computers, laptops, smartphones or personnel carriers to acquire images of the real object to be detected, or acquire images of the real object to be detected including the surrounding environment.

[0131] Step 2: Transmit the images acquired in Step 1 to the artificial intelligence system to complete the identification and recognition of the real objects to be detected;

[0132] Step 3: The detection module receives a real image, virtual image, or virtual representation of the real object to be detected from the real entity unit, as well as information or attributes related to the identity of the real object to be detected.

[0133] Step 4: The detection module transmits the real image, virtual image, or virtual representation of the real object to be detected and / or identity-related information or attributes received in Step 3 to the artificial intelligence system;

[0134] Step 5: Perform at least one of the following operations in real-time under real-world conditions: verify the operation of the AI ​​system; manage the operation of the AI ​​system in real-time under real-world conditions; monitor the operation of the AI ​​system in real-time under real-world conditions; correct the operation of the AI ​​system in real-time under real-world conditions; enhance the operation of the AI ​​system in real-time under real-world conditions; optimize the operation of the AI ​​system in real-time under real-world conditions.

[0135] The present invention also provides a device for detecting and identifying biological or non-biological entities that enables an artificial intelligence system to achieve self-correction and / or self-repair and / or self-generation and / or self-construction and / or self-optimization and / or self-training. The device includes at least one real entity unit and at least one detection module. The real entity unit is integrated into or embedded in the real object to be detected, or is associated with the real object to be detected, or partially or entirely replaces the real object to be detected.

[0136] The detection module is connected to the onboard local artificial intelligence system embedded in the detection module, or to the remote artificial intelligence system;

[0137] The artificial intelligence system is connected to one or more cameras and / or lidar and / or radar or any computer vision means and / or sound acquisition device, configured to capture images and / or shapes and / or sounds of a real object to be detected, with or without the environment in which the real object to be detected or identified by the artificial intelligence is located or active, and the cameras are visible light cameras and / or thermal imaging cameras.

[0138] This invention also provides a method for artificial intelligence self-correction and / or self-repair and / or self-generation and / or self-optimization and / or self-training, implemented using the aforementioned device for detecting and identifying biological or non-biological entities that enables artificial intelligence systems to achieve self-correction and / or self-repair and / or self-generation and / or self-construction and / or self-optimization and / or self-training, comprising the following steps:

[0139] Step 1: Using one or more cameras and / or lidar and / or radar and / or any computer vision device and / or sound acquisition device, acquire image and / or sound data of the real object to be detected, or the real object to be detected with its surrounding environment. The camera is a visible light camera and / or a thermal imaging camera. The device is configured to acquire the image, shape and / or sound of the real object to be detected. The acquisition process may or may not include the environment in which the real object is located or is active. The real object to be detected is detected or identified by artificial intelligence.

[0140] Step 2: Transmit the data or information collected in Step 1 to the artificial intelligence system for identification and recognition of the real object to be detected, whether or not it is attached to the environment in which the real object to be detected is located or active.

[0141] Step 3: The artificial intelligence system processes the data transmitted in Step 2 and outputs the processing results;

[0142] Step 4: The detection module receives the real image, virtual image or virtual representation of the real object to be detected, and / or information or attributes related to the identity of the real object to be detected emitted by the real entity unit. The detection module is associated with or connected to an artificial intelligence system with self-correction and / or self-repair and / or self-generation and / or self-construction and / or self-optimization and / or self-training functions.

[0143] Step 5: The detection module transmits the real image, virtual image or virtual representation of the real object to be detected received in Step 4, and / or the information or attributes related to the identity of the real object to be detected emitted by the real entity unit to the artificial intelligence system.

[0144] Step 6: Verify and / or control the operation of the artificial intelligence system in real-time under real-world conditions, and / or monitor the operation of the artificial intelligence system in real-time under real-world conditions, and / or correct the operation of the artificial intelligence system in real-time under real-world conditions, and / or enhance the operation of the artificial intelligence system in real-world conditions, and / or optimize the operation of the artificial intelligence system in real-world conditions; and / or verify or correct the results output by the artificial intelligence system in Step 3, thereby enabling the artificial intelligence system to achieve self-correction and / or self-repair and / or self-generation and / or self-construction and / or self-optimization and / or self-training.

[0145] Furthermore, steps 1 to 6 are performed either in chronological order or not in chronological order.

[0146] Furthermore, step 4 is performed simultaneously with steps 1 and / or 2 and / or 3. Attached Figure Description

[0147] Figure 1A schematic diagram showing the structure of a billboard replacing a physical unit of the device of the present invention;

[0148] Figure 2 A schematic diagram showing the structure of a road traffic sign replaced by a physical unit of the device of the present invention;

[0149] Figure 3 This is a schematic diagram showing the actual physical unit of the device of the present invention assembled in a shop.

[0150] Figure 4 This is a schematic diagram illustrating the implementation of the device of the present invention mounted on a truck and a passenger vehicle;

[0151] Figure 5 This is a schematic diagram illustrating the implementation of the device of the present invention mounted on a bicycle and a passenger vehicle;

[0152] Figure 6 This is a schematic diagram illustrating the implementation of the device of the present invention mounted on a passenger vehicle and a road for vehicle speed control;

[0153] Figure 7 This is a schematic diagram illustrating the implementation of the device of the present invention mounted on multiple passenger vehicles and on the road;

[0154] Figure 8 This is a schematic diagram illustrating the implementation of the device of the present invention mounted on multiple passenger vehicles and on the road. Detailed Implementation

[0155] Figure 1 An embodiment of overall replacement and modification is shown: the billboard 11 is completely replaced with a pole 13, and a real physical unit is mounted on the pole 13; or the billboard 11 is completely replaced with a tree 14, and a real physical unit 12 is mounted on the tree 14.

[0156] Figure 2 Another embodiment of overall replacement and modification is shown: replacing the road traffic sign 25 entirely with a pole 23, on which a real physical unit 22 is mounted; or replacing the road traffic sign 25 entirely with a tree 24, on which a real physical unit 22 is mounted.

[0157] Figure 3 An example of shop renovation is shown: embedding a real entity unit 32 into the signboard 36 of the shop 37; or using the real entity unit 32 to replace or completely replace the original signboard 36 of the shop 37.

[0158] Figure 4 An example of a detection application is shown: a semi-trailer truck 48 equipped with a physical unit 42 can be identified and detected by a passenger vehicle 49 equipped with a detection module 140. An example of information received by the detection module 140 on the passenger vehicle 49 is: there is a Ford 38-ton semi-trailer truck traveling in the same direction as this vehicle 100 meters ahead.

[0159] Figure 5 An example of non-motorized vehicle detection is shown: a cyclist 151 and his bicycle 152 are equipped with a physical unit 52, which can be detected and identified by a passenger vehicle 59 equipped with a detection module 150. An example of information received by the detection module 150 on the passenger vehicle 59 is: a young man is riding a Decathlon 2.5-inch wheel bicycle 400 meters ahead, traveling in the same direction as this vehicle.

[0160] Figure 6 An example of an application of the speeding detection method is shown: A passenger car 69 equipped with a physical unit is traveling on road segment 163; two detection modules 160a and 160b are deployed along this road segment, with a distance d between them. Both detection modules 160a and 160b are connected to a traffic violation data processing and ticket issuance center. The local traffic management department stipulates that the speed limit for section d on this road segment is VL. At time t1, vehicle 69 reaches detection module 160a, which receives information sent by the physical unit 62 onboard vehicle 69 and records time t1. An example of the received information is: red Citroën C5 sedan, license plate number 400VB75, owner Mr. DuPont, address 45 Rue de la Radar, Paris, France, postal code 75000.

[0161] At time t2, vehicle 69 approaches detection module 160b. This module receives information from the physical unit 62 mounted on vehicle 69 and records time t2. It also receives vehicle and owner information as described above. The system uploads the two timestamps, t1 and t2, along with the vehicle and owner information, to the data processing center. The data processing center calculates the time difference t2-t1. If the vehicle is speeding, the time difference t2-t1 is less than d / VL, and the system generates a speeding violation record and sends a penalty notice to Mr. DuPont's registered address.

[0162] Figure 7 An example of a road network application is shown: multiple passenger vehicles 79 equipped with detection modules 170 travel on the road network, and multiple real physical units 72 are deployed in a grid layout within the road network.

[0163] Figure 8 An aviation scenario example is shown: an aircraft 185 equipped with a detection module 180 flies over an area 186; multiple real entity units 2 that replace billboards are deployed in this area, and the real entity units push local business and tourism related advertising information to passengers (not shown) on the aircraft 185.

[0164] It should be understood that the apparatus and methods described above in conjunction with the accompanying drawings are only for disclosing the technical solutions of the present invention, and are not intended to limit the present invention. Those skilled in the art can make various adjustments, modifications, and optimizations to the above embodiments without departing from the scope of protection of the present invention.

Claims

1. A device for detecting and identifying biological or non-biological entities, characterized in that, include: At least one detectable real entity unit, said real entity unit being mounted on a real object to be detected and identified; and at least one detection module, the detection module being configured on another component, for detecting the real entity unit; The real entity unit broadcasts the real image, virtual image or virtual image of the real object to be detected, as well as information or attributes related to the identity of the real object to be detected, in a one-way, non-interactive and cyclical manner. The real entity unit has a unique digital identity and includes: A storage and / or digital recording device for storing real and / or virtual images of the real object to be detected, and / or information or attributes related to the identity of the real object to be detected; An internal power supply device is used to supply power to the actual physical unit; A remote receiving device is used to remotely receive information related to the real object to be detected, as well as updated data of the real image or virtual image of the real object to be detected. The acquisition device is used to acquire real images of the real object to be detected, the real images being with or without relevant environmental information. A broadcasting device for cyclically broadcasting data to one or more mobile or fixed detection modules; Each of the detection modules is equipped with: A receiving device is configured to receive image and / or information data broadcast by the real entity unit. A temporary storage device for temporarily storing the data. Display and / or listening devices for displaying and / or listening to the data. A processing device for processing the data.

2. The apparatus according to claim 1, characterized in that, The broadcast was a television broadcast.

3. The apparatus according to claim 1, characterized in that, The broadcast is a radio broadcast used to transmit audio narration recordings of the real object to be detected, as well as information associated with the identity of the real object to be detected.

4. The apparatus according to claim 1, characterized in that, The audio narration of the real object to be detected is information or attribute associated with the real object to be detected.

5. The apparatus according to claim 1, characterized in that, The information or attributes broadcast that are associated with the real object to be detected include the object's position, orientation, direction of motion, and speed.

6. The apparatus according to claim 1, characterized in that, The storage capacity of the storage device and / or digital recording device is in the range of 0.1 megabytes to 10 terabytes and above.

7. The apparatus according to claim 1, characterized in that, The internal power supply device is rechargeable.

8. The apparatus according to claim 1, characterized in that, The internal power supply device is connected to an external charging device, which includes a solar panel, a wind turbine, a magnetic field device, a piezoelectric device, or a power grid.

9. The apparatus according to claim 1, characterized in that, The remote receiving device has a receiving range of 500 meters to 100,000 kilometers.

10. The apparatus according to claim 1, characterized in that, It also includes a conversion device for converting the real image of the real object to be detected into a virtual image.

11. The apparatus according to claim 1, characterized in that, The television broadcast and / or radio broadcast loops are executed in cycles at a frequency of 1 to 120 seconds, with a transmission distance of 500 to 20,000 meters, and the data transmission rate of the television broadcast and / or radio broadcast is 100 kilobytes / second to 10 gigabytes / second and above.

12. The apparatus according to claim 1, characterized in that, The detection module is embedded or integrated into wheeled machines, flying machines, autonomous or semi-autonomous wheeled machines, autonomous or semi-autonomous flying machines, desktop computers, or smartphones, or is carried by personnel.

13. The apparatus according to claim 1, characterized in that, The detection device for detecting the precise position, posture, direction of motion, and speed of the real object to be detected consists of a GPS sensor, an inertial gyroscope sensor, and an accelerometer.

14. The apparatus according to claim 1, characterized in that, The detection module and / or physical unit are integrated into backpacks, handbags, suitcases, bicycle helmets, and motorcycle helmets for human use.

15. The apparatus according to claim 1, characterized in that, The detection module and / or physical entity unit are composed of a smartphone, which is modified to enable the device to have additional functions. These additional functions are controlled by a dedicated application, and the driving and safety programs of the wheeled vehicle are compatible with the smartphone.

16. The apparatus according to claim 1, characterized in that, It also includes a detection device, which is used to detect the precise position, posture, direction of motion, and speed of motion of the real object to be detected.

17. The apparatus according to claim 1, characterized in that, The data processing device of the detection module communicates with the guidance and control systems of flying and wheeled machines.

18. The apparatus according to claim 1, characterized in that, The detection module also includes a timing device, which is used to determine the connection duration between the detection module and the real entity unit.

19. A method for detecting and identifying biological or non-biological entities, characterized in that, Includes the following steps: The real object to be detected is modified to make it a detectable target. The real object to be detected can be fixed or mobile, inanimate or living. The modification method is to integrate or embed real entity units into the real object to be detected, or establish an association with the real object to be detected, or use the real entity units to partially or completely replace the real object to be detected. The real entity unit broadcasts the real image, virtual image or virtual representation corresponding to the real object type to be detected, as well as information or attributes related to the identity of the real object to be detected, in a one-way transmission, no two-way interactive communication and cyclic transmission television broadcast manner, and / or broadcasts the audio explanation voice of the real object type to be detected and the identity-related information or attributes in a cyclic transmission radio broadcast manner. The types of real objects to be detected include pedestrians, animals, bicycles, motorcycles, wheeled transport equipment, aircraft, billboards, pre-advertisement signs, advertising signs, and road signs; the identity-related information or attributes include object type, object properties, object name, object content, and object size. The real entity unit has a unique digital identity and includes: A storage and / or digital recording device is used to store real images, virtual images, and information or attributes related to the identity of the real objects to be detected for various types of real objects, and / or to store and save audio narration voice of the type of real object to be detected and information or attributes related to its identity. The storage capacity of the storage device and / or digital recording device is in the range of 0.1 megabytes to 10 terabytes and above. An internal power supply device is used to power the real physical unit. The internal power supply device may be rechargeable or non-rechargeable and may be connected to an external charging device, which may include a solar panel, a wind turbine, a magnetic field device, a piezoelectric device, or a power grid. A remote receiving device is used to remotely receive information or attributes related to the identity of a real object to be detected, or updated data of a real or virtual image, via a digital wireless link or wired link, with a receiving range of 500 meters to 100,000 kilometers. The acquisition device is used to acquire real images of a real object to be detected, with or without a relevant environment, and the conversion device is used to convert the real images of the real object to be detected into virtual images to ensure the confidentiality of the real images of the real object to be detected. The detection device is used to acquire the position, posture, direction of motion, and speed of the real object to be detected. A broadcasting device is used to broadcast, in a unidirectional, non-bidirectional interactive communication, and cyclical manner, real images, virtual images, or virtual representations of a real object to be detected, as well as information or attributes related to the identity of the real object to be detected, on television, with or without the object's position, posture, direction of motion, and speed; and / or radio broadcasts audio of the real object to be detected, along with identity-related information or attributes, with or without the object's position, posture, direction of motion, and speed; the execution cycle of the television broadcasting and / or radio broadcasting loop is from 1 second to 120 seconds, the transmission distance is from 500 meters to 20,000 meters, and the data transmission rate of the television broadcasting and / or radio broadcasting is from 100 kilobytes / second to 10 gigabytes / second and above; The system broadcasts a real image, virtual image, or virtual representation of the real object to be detected, along with identity-related information or attributes, to one or more detection modules via television in a one-way transmission, without bidirectional interactive communication and in a loop. The television broadcast content may or may not include the object's position, posture, direction of motion, and speed. Alternatively, the system broadcasts an audio narration of the real object to be detected, along with identity-related information or attributes, to one or more detection modules via radio in a one-way transmission, without bidirectional interactive communication and in a loop. The radio broadcast content may or may not include the object's position, posture, direction of motion, and speed. The detection module is embedded or integrated into portable terminals such as wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines, autonomous / semi-autonomous flying machines, desktop computers, and smartphones, or is carried by the human body. Each detection module is equipped with: A receiving device is used to receive a real or virtual image of a real object to be detected with information or attributes related to its identity, and / or to receive audio narration of the real object to be detected and information or attributes related to the identity of the real object to be detected. A storage device for temporarily storing real or virtual images of a real object to be detected with identity-related information or attributes, and / or audio narration of the real object to be detected and information or attributes associated with the identity of the real object to be detected. A display or projection apparatus, configured to display or project a real image or a virtual image of a to-be-detected real object attached with identity-related information or attributes, and / or a listening apparatus, configured to play the audio commentary voice of the to-be-detected real object and the information or attributes related to the identity of the to-be-detected real object; A processing apparatus, configured to perform arithmetic processing on the real image and the virtual image of the to-be-detected real object attached with identity-related information or attributes; The detection module is connected to one or more real entity units at the same time, with a single connection duration ranging from 2 seconds to 10 minutes; the detection module receives and displays on its own built-in display apparatus the real image or virtual image of the to-be-detected real object as well as the identity-related information or attributes, wherein the information received and displayed by the detection module may or may not include the position of the object, the pose of the object, the movement direction, and the movement speed; and / or transmits the type of the identified real object and the identity-related information or attributes to the control and guidance system of a semi-autonomous or autonomous wheeled / flying machine, wherein the transmitted information may or may not include the position of the object, the pose of the object, the movement direction, and the movement speed; and / or the detection module is connected to one or more real entity units at the same time, with a single connection duration ranging from 2 seconds to 10 minutes; the detection module receives the audio commentary voice of the to-be-detected real object and the identity-related information or attributes, and outputs the audio commentary voice to the listening apparatus supporting the detection module; and / or transmits the type of the identified real object and the identity-related information or attributes to an audio information processing system, so as to realize the control and guidance of the semi-autonomous / autonomous wheeled machine and flying machine; the audio commentary voice and the identity-related information or attributes may or may not be attached with information of the physical object position, the object pose, the movement direction, and the movement speed.

20. The method according to claim 19, characterized in that, When adopting a wholly-replaced entity transformation solution for billboards, advertisement signboards, and advertisement preview signs, the original entity is replaced with a replacement entity, wherein the replacement entity is a natural tree, an artificial tree, a natural flower, or an artificial flower, and the real entity unit is integrated into the replacement entity.

21. The method according to claim 19, characterized in that, In the scenarios of guidance, autonomous guidance and control of autonomous or semi-autonomous devices, the information acquired by the detection module from the real entity unit can drive the device manipulation mechanism to perform actions, where the actions include steering, acceleration, and braking.

22. The method according to claim 19, characterized in that, The system further includes configuring a detection apparatus for the real entity unit, wherein the detection apparatus is configured to detect the precise position of the to-be-detected real object, the pose of the object, the movement direction of the object, and the movement speed of the object.

23. The method according to claim 19, characterized in that, The image processing apparatus of the detection module communicates with the guidance and control system; and / or is equipped with a speech recognition apparatus carrying artificial intelligence, and an audio processing apparatus for processing received audio information, wherein the speech recognition apparatus and the audio processing apparatus establish communication with the guidance and control system of the wheeled machine and the flying machine.

24. The method according to claim 19, characterized in that, The system further includes configuring a timing apparatus for the detection module, wherein the timing apparatus is configured to measure the connection duration between the detection module and the real entity unit.

25. The method according to claim 19, characterized in that, It also includes deploying detection modules at intervals along a section of road. The detection modules are deployed along the road at certain intervals, and the speed of the vehicle on the road section is calculated by measuring the travel time of a vehicle carrying a real physical unit through two detection modules.

26. The method according to claim 19, characterized in that... The method further includes the following steps: Identify all possible motion postures of the real object to be detected. For each possible motion posture of the real object to be detected, a corresponding audio description, visual description, and / or encoding is matched. One or more sensors are configured for each possible motion posture of the real object to be detected. These sensors are used to detect or acquire the motion state of the entity. A real entity unit is configured for the real object to be detected. The real entity unit has built-in sensors corresponding to various motion actions that can be performed by all the identified entities. The actual motion of the real object being detected is collected by various sensors. Using the broadcasting device built into the real entity unit, descriptive information and / or codes matching the motion of the real object to be detected collected by the sensor are broadcast to the detection module. The broadcasting device is configured to transmit the corresponding codes or descriptive information of the motion of the real object to be detected in a one-way, non-interactive, looping television broadcasting mode and / or radio broadcasting mode. The television broadcasts and / or radio broadcasts have a cycle period of 1 to 120 seconds, and the long-distance broadcast coverage range is 500 meters to 20,000 meters. The transmission rate of the television broadcasts and / or radio broadcasts is 100 kilobytes per second to 10 gigabytes per second and above. The detection module is equipped with a receiving device, which is used to receive audio or visual descriptions and / or codes associated with the motion of the real object to be detected. The description information is amplified by the sound playback device on the detection module, or displayed on the display device on the detection module, or the code is transmitted to the computing device, which retrieves a database containing codes corresponding to various feasible movements to complete the encoding and decoding or matching and association with the movement. The description information and / or code corresponding to the movement can be used to implement manual or autonomous automatic control of the motion control mechanism of the real object to be detected.

27. The method according to claim 19, characterized in that, The detection module is in the same environment as the real entity to be detected, or moves and operates in the same environment.

28. The method according to claim 27, characterized in that, The real object to be detected and the environment in which the detection module is located or operates are aircraft, land-based wheeled vehicles, water vehicles, or buildings.

29. The method according to claim 27 or 28, characterized in that, The broadcast transmission distance of the real image, virtual image, or virtual image of the real object to be detected, as well as the information or attributes related to the identity of the real object to be detected, is limited by the spatial boundary of the environment.

30. The method according to any one of claims 27 to 29, characterized in that, The broadcast transmission distance for the real image, virtual image, or virtual representation of the real object to be detected, as well as information or attributes related to the identity of the real object to be detected, is from 1 meter to 500 meters.

31. An apparatus for automatically training artificial intelligence using the detection device of claim 1 through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning, characterized in that: Wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines, autonomous / semi-autonomous flying machines, desktop computers, smartphones, or personnel equipped with the aforementioned detection module are all equipped with: An onboard local artificial intelligence system, or a communication connection established with a remote artificial intelligence system; One or more cameras and / or lidar or any computer vision acquisition device for acquiring images and contours of the real object to be detected, wherein the camera is a visible light camera and / or a thermal imaging camera. One or more detection modules are connected to an onboard local artificial intelligence system mounted on an autonomous or semi-autonomous wheeled or flying machine, a desktop or smartphone-like portable terminal, or a human body carrier, or to a remote artificial intelligence system. The autonomous or semi-autonomous wheeled or flying machine, desktop or smartphone-like portable terminal, and human body carrier are all connected to the onboard local artificial intelligence system or the remote artificial intelligence system.

32. A method for automatically training artificial intelligence through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning using the detection device of claim 1, the detection method of claim 19, and the automatic training device of claim 31, characterized in that, Includes the following steps: Step 1: Use cameras and / or lidar and / or other computer vision acquisition devices mounted on wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines and flying machines, desktop computers, laptops, smartphones or personnel carriers to acquire images of the real object to be detected, or acquire images of the real object to be detected including the surrounding environment. Step 2: Transmit the images acquired in Step 1 to the artificial intelligence system to complete the identification and recognition of the real objects to be detected; Step 3: The detection module receives a real image, virtual image, or virtual representation of the real object to be detected from the real entity unit, as well as information or attributes related to the identity of the real object to be detected. Step 4: The detection module transmits the real image, virtual image, or virtual representation of the real object to be detected and / or identity-related information or attributes received in Step 3 to the artificial intelligence system; Step 5: Verify, correct, or optimize the real object recognition results output by the artificial intelligence system in Step 2, and / or automatically train the artificial intelligence through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning.

33. A method for enhancing the performance of artificial intelligence and / or optimizing its operational effects using the detection device of claim 1, the detection method of claim 19, and the automatic training device of claim 31, characterized in that, Includes the following steps: Step 1: Use cameras and / or lidar and / or other computer vision acquisition devices mounted on wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines and flying machines, desktop computers, laptops, smartphones or personnel carriers to acquire images of the real object to be detected, or acquire images of the real object to be detected including the surrounding environment. Step 2: Transmit the images acquired in Step 1 to the artificial intelligence system to complete the identification and recognition of the real objects to be detected; Step 3: The detection module receives a real image, virtual image, or virtual representation of the real object to be detected from the real entity unit, as well as information or attributes related to the identity of the real object to be detected. Step 4: The detection module transmits the real image, virtual image, or virtual representation of the real object to be detected and / or identity-related information or attributes received in Step 3 to the artificial intelligence system; Step 5: Verify, correct, or optimize the recognition results of the real objects to be detected output by the artificial intelligence system in Step 2.

34. The method for automatically training artificial intelligence through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning, as described in claim 32 or 33, and / or the method for enhancing the performance of artificial intelligence and / or optimizing the operational effect of artificial intelligence, characterized in that... Steps 1 through 5 can be performed in chronological order or not.

35. The method for automatically training artificial intelligence through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning according to claim 32 or 33, and / or the method for enhancing the performance of artificial intelligence and / or optimizing the effect of artificial intelligence operation, characterized in that, Step 3 is executed simultaneously with Step 1 and / or Step 2.

36. The method for automatically training artificial intelligence through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning, as described in claim 32 or 33, and / or the method for enhancing the performance of artificial intelligence and / or optimizing the operational effect of artificial intelligence, characterized in that... Step 3 is executed simultaneously with Step 1 and / or Step 2.

37. The method according to any one of claims 19 to 30, characterized in that, The broadcast of the real image, virtual image or virtual image of the real object to be detected, as well as the information or attributes associated with the identity of the real object to be detected, is transmitted using 4G broadcast, 5G broadcast or 6G broadcast.

38. A method for automatically training artificial intelligence in real-time and real-world operating conditions through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning using the detection device of claim 1, the detection method of claim 19, the automatic training device of claim 31, and the method of any one of claims 32 to 35, characterized in that... Includes the following steps: Step 1: Use cameras and / or lidar and / or other computer vision devices mounted on wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines and flying machines, desktop computers, laptops, smartphones or personnel carriers to acquire images of the real object to be detected, or acquire images of the real object to be detected including the surrounding environment. Step 2: Transmit the images acquired in Step 1 to the artificial intelligence system to complete the identification and recognition of the real objects to be detected; Step 3: The detection module receives a real image, virtual image, or virtual representation of the real object to be detected from the real entity unit, as well as information or attributes related to the identity of the real object to be detected. Step 4: The detection module transmits the real image, virtual image, or virtual representation of the real object to be detected and / or identity-related information or attributes received in Step 3 to the artificial intelligence system; Step 5: Under real-time and real-world conditions, verify, correct, or optimize the real object recognition results output by the artificial intelligence system in Step 2, and / or automatically train the artificial intelligence through supervised learning and / or unsupervised learning and / or reinforcement learning and / or deep learning.

39. A method for controlling and / or supervising and / or correcting and / or enhancing the operation of an artificial intelligence system and / or optimizing the operation of an artificial intelligence system under real-time and real-world operating conditions using the detection device of claim 1, the detection method of claim 19, the automatic training device of claim 31, and any one of claims 32 to 35, characterized in that, Includes the following steps: Step 1: Use cameras and / or lidar and / or other computer vision devices mounted on wheeled machines, flying machines, autonomous / semi-autonomous wheeled machines and flying machines, desktop computers, laptops, smartphones or personnel carriers to acquire images of the real object to be detected, or acquire images of the real object to be detected including the surrounding environment. Step 2: Transmit the images acquired in Step 1 to the artificial intelligence system to complete the identification and recognition of the real objects to be detected; Step 3: The detection module receives a real image, virtual image, or virtual representation of the real object to be detected from the real entity unit, as well as information or attributes related to the identity of the real object to be detected. Step 4: The detection module transmits the real image, virtual image, or virtual representation of the real object to be detected and / or identity-related information or attributes received in Step 3 to the artificial intelligence system; Step 5: Perform at least one of the following operations in real-time under real-world conditions: verify the operation of the AI ​​system; manage the operation of the AI ​​system in real-time under real-world conditions; monitor the operation of the AI ​​system in real-time under real-world conditions; correct the operation of the AI ​​system in real-time under real-world conditions; enhance the operation of the AI ​​system in real-time under real-world conditions; optimize the operation of the AI ​​system in real-time under real-world conditions.

40. The apparatus for detecting and identifying biological or non-biological entities according to any one of claims 1 to 18, allowing artificial intelligence to self-correct and / or self-repair and / or self-generate and / or self-build and / or self-optimize and / or self-train, characterized in that: The device includes at least one detectable real entity unit and at least one detection module. The real entity unit is associated with, integrated with, embedded in, or partially replaces or does not replace the real object to be detected and identified. The detection module establishes a connection with the embedded onboard local artificial intelligence system or with a remote artificial intelligence system. The artificial intelligence system is connected to one or more cameras and / or lidar and / or radar or any computer vision device and / or sound acquisition device, configured to acquire images and / or shapes and / or sounds of real objects to be detected, with or without the environment in which the real objects to be detected or identified by artificial intelligence are located or active, and the cameras are visible light and / or thermal imaging cameras.

41. A method using the apparatus of claim 40 that allows for self-correction and / or self-repair and / or self-generation and / or self-construction and / or self-optimization and / or self-training of artificial intelligence, characterized in that, Includes the following steps: Step 1: Using one or more cameras and / or lidar and / or radar and / or any computer vision device and / or sound acquisition device, acquire image and / or sound data of the real object to be detected, or the real object to be detected with its surrounding environment. The camera is a visible light camera and / or a thermal imaging camera. The device is configured to acquire the image, shape and / or sound of the real object to be detected. The acquisition process may or may not include the environment in which the real object is located or is active. The real object to be detected is detected or identified by artificial intelligence. Step 2: Transmit the data or information collected in Step 1 to the artificial intelligence system for identification and recognition of the real object to be detected, whether or not it is attached to the environment in which the real object to be detected is located or active. Step 3: The artificial intelligence system processes the data transmitted in Step 2 and outputs the processing results; Step 4: The detection module receives the real image, virtual image or virtual representation of the real object to be detected, and / or information or attributes related to the identity of the real object to be detected emitted by the real entity unit. The detection module is associated with or connected to an artificial intelligence system with self-correction and / or self-repair and / or self-generation and / or self-construction and / or self-optimization and / or self-training functions. Step 5: The detection module transmits the real image, virtual image or virtual representation of the real object to be detected received in Step 4, and / or the information or attributes related to the identity of the real object to be detected emitted by the real entity unit to the artificial intelligence system. Step 6: Verify and / or control the operation of the artificial intelligence system in real-time under real-world conditions, and / or monitor the operation of the artificial intelligence system in real-time under real-world conditions, and / or correct the operation of the artificial intelligence system in real-time under real-world conditions, and / or enhance the operation of the artificial intelligence system in real-world conditions, and / or optimize the operation of the artificial intelligence system in real-world conditions; and / or verify or correct the results output by the artificial intelligence system in Step 3, thereby enabling the artificial intelligence system to achieve self-correction and / or self-repair and / or self-generation and / or self-construction and / or self-optimization and / or self-training.

42. The method for enabling artificial intelligence to self-correct and / or self-repair and / or self-generate and / or self-build and / or self-optimize and / or self-train according to claim 41, characterized in that, Steps 1 through 6 can be performed in chronological order or not.

43. The method for enabling artificial intelligence to self-correct and / or self-repair and / or self-generate and / or self-build and / or self-optimize and / or self-train according to claim 41, characterized in that, Step 4 can be performed simultaneously with Step 1 and / or Step 2 and / or Step 3.

Images