System for smart safety detection and worker behavior monitoring

Abstract

The present invention discloses a technical concept comprising: an imaging unit attached to any position of a vehicle used at a specific site, which captures a specific site from any position and detects visual information; an alarm unit attached to the rear of the vehicle, which transmits a warning from the rear of the vehicle; a display unit provided to the driver's seat of the vehicle, which provides the driver with information about a specific site captured by the imaging unit; and a signalman alarm unit that warns a signalman performing safety control near the vehicle of the approach of the vehicle.

Description

Technology Field

[0001] The present invention is intended to prevent safety accidents at vehicle or heavy equipment work sites. More specifically, it relates to a heavy equipment collision warning system that simplifies the installation of safety equipment for owner-operator vehicles participating in construction and industrial sites to reduce costs and time, prevents collision accidents by detecting the human body in real time through AI video analysis by attaching a magnetic-type main body to the rear of the heavy equipment and a monitor to the driver's seat. Background Technology

[0003] Safety at industrial and construction sites is emerging as a critical issue for companies and the government.

[0004] Numerous industrial accidents occur worldwide every year, and these accidents can lead not only to loss of life but also to economic losses, legal liability, and reduced productivity for companies.

[0005] Accordingly, the introduction of smart safety monitoring systems to ensure worker safety and prevent accidents is becoming increasingly widespread.

[0006] The global safety monitoring system market is estimated to be between $6 billion and $8 billion as of 2023 and is projected to grow rapidly at a compound annual growth rate (CAGR) of 8 to 12% until 2028.

[0007] The increasing adoption of smart factories, the expansion of automation in industries such as construction, manufacturing, and logistics, stricter government regulations, and the spread of ESG (Environmental, Social, and Governance) management are becoming major growth drivers for the safety monitoring market.

[0008] Depending on the country, government regulations and legal requirements are increasing, and agencies such as the U.S. Occupational Safety and Health Administration (OSHA) and the European EU-OSHA are tightening safety regulations, imposing fines or taking legal action against companies that fail to comply.

[0009] In addition, South Korea has further strengthened safety obligations in the workplace through measures such as the amendment of the Industrial Safety and Health Act (introduction of the Serious Accidents Punishment Act).

[0010] Recently, with the advancement of AI, IoT, machine learning, and cloud-based technologies, automated AI surveillance systems are gaining attention to replace existing manual surveillance systems, and various smart safety systems are emerging due to technological progress.

[0011] For example, technologies that analyze workers' behavior in real time through AI-based video analysis and detect biosignals using wearable devices are rapidly developing, and technological attempts are being made.

[0012] As a technical attempt regarding the subject matter, there exists a prior patent document titled "Intelligent Heavy Equipment Safety Remote Control System Using an Access Alarm Module Based on WB Ultra-Wideband Communication (Registration No. 10-2072380, hereinafter referred to as Patent Document 1)."

[0013] Patent Document 1 relates to an intelligent heavy equipment safety remote control system using an approach alarm module based on UWB ultra-wideband communication, and more specifically, to an approach alarm module comprising a worker UWB tag attached to a worker, a heavy equipment UWB tag attached to heavy equipment, a facility UWB anchor and UWB tag attached to a fixed facility, and an indicator attached to heavy equipment that generates an alarm when the worker UWB tag, the heavy equipment UWB tag, and the facility UWB tag approach within a certain distance; a remote maintenance control unit (RMCU) for communicating with the worker UWB tag, the heavy equipment UWB tag, the facility UWB anchor and UWB tag, and the indicator to output information regarding alarm events, the location and status of heavy equipment; The present invention relates to an intelligent heavy equipment safety remote control system using an access alarm module based on UWB ultra-wideband communication, comprising a manager terminal that communicates with a remote maintenance control device and outputs a display screen for controlling and monitoring alarm events, the location and status of heavy equipment, and the safety of workers, thereby enabling integrated monitoring of the location and status of heavy equipment and the safety of workers at a construction or industrial site.

[0014] Another related prior art patent is the "Industrial Site Disaster Prevention and Safety Integrated Monitoring System Equipped with Intelligent Detection Sensors (Publication No. 10-2023-0033778, hereinafter referred to as Patent Document 2)."

[0015] Patent Document 2 relates to an integrated monitoring system for industrial site disaster prevention and safety equipped with an intelligent sensing sensor, comprising a sensing unit equipped with at least one intelligent IoT sensor, a wireless signal relay that transmits the signal of the sensing unit to a receiving unit, an IoT gateway that receives the signal from the wireless signal relay and converts it into another communication protocol, and a receiving unit equipped with an artificial intelligence DB that stores and analyzes data collected through the IoT gateway.

[0016] Likewise, "LTE / 5G communication and IoT-based intelligent control monitoring system for safety response to prevent construction equipment accidents (Publication No. 10-2021-0025311, hereinafter referred to as Patent Document 3)" also corresponds to a related prior art patent document.

[0017] Patent Document 3 relates to an intelligent control monitoring system based on LTE / 5G communication and IoT for safety response to prevent accidents involving workers and construction equipment, which enables effective response to accident prevention involving workers and construction equipment using LTE / 5G communication. The present invention comprises: a control center server that sends a report command signal to an LTE / 5G network via LTE / 5G communication to report worker status information, construction equipment status information, and construction site status information; an LTE / 5G network that sends a report command signal received from the control center server to a gateway via LTE / 5G communication to report worker status information, construction equipment status information, and construction site status information; and a gateway that sends a report command signal received from the LTE / 5G network to an IoT terminal installed on one side of construction equipment located at the construction site via LTE / 5G communication to report worker status information, construction equipment status information, and construction site status information.

[0018] These prior art patent documents lack technology for video analysis and precise behavior detection, relying on sensor and network technologies to detect safety status and provide notifications to managers, and lack access detection and warning systems for hazardous areas, relying on individual monitoring of worker safety. Prior art literature

[0020] Registration No. 10-2072380 Publication No. 10-2023-0033778 Publication No. 10-2021-0025311 The problem to be solved

[0021] The heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention has been devised to solve the conventional problems described above and presents the following problem to be solved.

[0022] First, it enables effective detection and warning of collision accidents involving workers that may occur in blind spots around vehicles or heavy equipment.

[0023] Second, we aim to minimize false alarms and improve system reliability by accurately identifying only human bodies at the work site.

[0024] Third, it is designed with neodymium (permanent magnet) so that it can be easily attached to a vehicle, and aims to provide an installation method that can drastically reduce time and costs at short-term work sites.

[0025] Fourth, we intend to improve the safety of signalmen by providing an alarm bell to signalmen who are most exposed to danger by controlling traffic and personnel near heavy equipment.

[0026] The problems solved by the present invention are not limited to those mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the description below. means of solving the problem

[0028] The heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention has the following means for solving the problem to be solved above.

[0029] A heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention may be characterized by comprising: an imaging unit attached to any location of a vehicle used at a predetermined site, which captures the predetermined site from said arbitrary location and detects visual information; an alarm unit attached to the rear of said vehicle, which transmits a warning from the rear of said vehicle; and a display unit provided to the driver's seat of said vehicle, which provides the driver with information about the predetermined site captured by said imaging unit.

[0030] A heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to the present invention may further include a storage unit that receives information of a predetermined site captured through the imaging unit and stores information of the predetermined site.

[0031] The heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention may further include a signalman warning unit that warns a signalman of the approach of the vehicle.

[0032] The imaging unit of the heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to the present invention may be characterized by comprising: a shooting unit that captures a predetermined scene from the arbitrary location of the vehicle and captures the shooting information; and a biometric information detection unit that detects biometric information from the shooting information captured at the predetermined scene.

[0033] The imaging unit of the heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to the present invention may be characterized by including a boundary adjustment unit that freely adjusts the recognition range for a preset boundary.

[0034] The boundary adjustment unit of the heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to the present invention may be characterized by expanding or reducing the boundary size according to changes in the work environment.

[0035] The display unit of the heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention may be characterized by including: a video receiving unit that networks with the imaging unit and receives the field video transmitted from the imaging unit; and a display unit that displays the field video received from the video receiving unit to the driver.

[0036] The alarm unit of the heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention may be characterized by including a warning notification unit that provides notification to the driver, the pedestrian, and the signalman controlling safety when an abnormality occurs regarding preset safety information.

[0037] The notification unit of the heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention may be characterized by providing a warning to the pedestrian, the driver, and the signalman controlling safety when the captured information exceeds a certain level with respect to the preset safety information.

[0038] The signalman alarm unit of the heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention may be characterized by including an alarm device portion that is provided to the signalman as a wearable device and transmits a warning signal to the signalman.

[0039] The imaging unit of the heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention may be characterized by detecting the surrounding environment of the vehicle through a plurality of shooting units that photograph the work site, combining the shooting sites of the plurality of shooting units to determine abnormal movement regarding the movement of the pedestrian, and providing visual information from the vehicle to the display unit. Effects of the invention

[0041] Figure 4 shows a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention, which has the configuration as described above. The heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to the present invention provides the following effects.

[0042] First, we aim to ensure the life and safety of workers by accurately recognizing only biometric information at the work site to effectively prevent collision accidents between vehicles or heavy equipment and workers.

[0043] Second, the organic operation between the imaging unit, the display unit, and the alarm unit provides real-time warnings to the driver and operator, enabling immediate prevention of accidents.

[0044] Third, we aim to enhance work safety and maximize work efficiency through simple installation and intuitive usage.

[0045] Fourth, the aim is to provide an effect that can contribute to the analysis of the cause of accidents and the resolution of legal disputes by storing the video captured by the imaging unit and utilizing it as legal evidence in the event of an accident.

[0046] Fifth, by providing an alert to the signalman through the signalman alarm unit, the risk of accidents near heavy equipment is prevented, and the safety of the signalman is effectively ensured.

[0047] The effects of the present invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art from the description below. Brief explanation of the drawing

[0049] FIG. 1 is a conceptual diagram of a heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to one embodiment of the present invention. FIG. 2 illustrates an alarm unit, a display unit, and a signalman alarm unit connected via a network in a heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to an embodiment of the present invention. FIG. 3 illustrates that an imaging unit, an alarm unit, and a display unit of a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention are attached to heavy equipment via magnetism. FIG. 4 illustrates a heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to an embodiment of the present invention, wherein a signalman warning unit is provided to a signalman to provide a notification. Figure 5 illustrates that biometric information is detected through a biometric information detection unit in a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention. FIG. 6 illustrates a heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to an embodiment of the present invention, providing a danger warning to a pedestrian through a warning notification unit. FIGS. 7 and 8 illustrate a boundary adjustment unit in a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention, wherein a boundary is adjusted and set to guard a danger zone around a vehicle. FIG. 9 illustrates a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention, in which a predetermined scene is captured from various angles through a plurality of imaging units. FIG. 10 is a block diagram illustrating each unit of a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to one embodiment of the present invention. FIG. 11 is a block diagram illustrating the sub-configuration of an imaging unit in a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention. FIG. 12 is a block diagram illustrating the sub-configuration of a display unit in a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention. FIG. 13 is a conceptual diagram illustrating a method for storing information on biometric information stored in a storage unit of a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention. Specific details for implementing the invention

[0050] The heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention may be subject to various modifications and may have various embodiments. Specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, and it should be understood that it includes all modifications, equivalents, and substitutions that fall within the technical spirit and scope of the present invention.

[0051] FIG. 1 is a conceptual diagram of a heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to an embodiment of the present invention. FIG. 2 illustrates that an alarm unit, a display unit, and a signalman warning unit are connected via a network within a heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to an embodiment of the present invention. FIG. 3 illustrates that an imaging unit, an alarm unit, and a display unit are attached to heavy equipment via magnetism within a heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to an embodiment of the present invention. FIG. 4 illustrates that a signalman warning unit is provided to a signalman to provide a notification within a heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to an embodiment of the present invention. FIG. 5 illustrates that biometric information is detected through a biometric information detection unit within a heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to an embodiment of the present invention. FIG. 6 illustrates that a danger warning is provided to a pedestrian through a warning notification unit within a heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to an embodiment of the present invention. FIGS. 7 and 8 illustrate a boundary adjustment unit in which a boundary is adjusted and set to guard a danger zone around a vehicle in a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention. FIG. 9 illustrates a predetermined scene being captured from various angles through a plurality of imaging units in a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention. FIG. 10 is a block diagram illustrating each unit of a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention.FIG. 11 is a block diagram illustrating the sub-configuration of an imaging unit in a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention. FIG. 12 is a block diagram illustrating the sub-configuration of a display unit in a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention. FIG. 13 is a conceptual diagram illustrating a method for storing information regarding biometric information stored in a storage unit in a heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to an embodiment of the present invention.

[0052] The heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention is a system provided to a work environment that detects biometric information during the work environment, monitors the safety status of pedestrians through the biometric information, and provides notifications.

[0053] In the case of pedestrians, depending on the work site, they may include workers within the work site such as supervisors, managers, and mechanics, visitors outside the work site, and general pedestrians, movers, or visitors in the vicinity of the work site.

[0054] The heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention is configured to include an imaging unit (100), an alarm unit (200), and a display unit (300).

[0055] The imaging unit (100) is configured to be attached to any location of a vehicle used at a specific site, and to photograph the specific site from any location and detect visual information.

[0056] The imaging unit (100) captures a work environment corresponding to a predetermined site and simultaneously detects biometric information to collect data corresponding to visual information in real time.

[0057] In the case of the imaging unit (100), the main body can be easily attached to any location on the vehicle as shown in FIGS. 1 and 2, as the main body is magnetic.

[0058] In the case of such imaging units (100), they can be attached to a desired location on a vehicle and can be provided in multiple units.

[0059] In the case of the imaging unit (100), power can be supplied to the camera using the power from the camera cigarette lighter socket.

[0060] In the case of the alarm unit (200), it is attached to the rear of the vehicle and emits a warning from the rear of the vehicle.

[0061] The alarm unit (200) can provide danger to pedestrians and drivers through an LED warning light and a warning sound.

[0062] In the case of the alarm unit (200), when a pedestrian approaches the rear of the vehicle or a danger zone, it immediately activates a warning sound and a warning light to allow the pedestrian or driver to recognize the danger and respond.

[0063] In addition, the alarm unit (200) can provide a visual warning through a warning light along with a warning sound in noisy environments.

[0064] In the case of the display unit (300), it is provided in the driver's seat of the vehicle and provides the driver with information about a specific scene captured by the shooting unit (100).

[0065] The display unit (300) displays the visual information transmitted from the imaging unit (100) in real time.

[0066] Additionally, the display unit (300) may also highlight a visual warning message or a dangerous location on the display when it detects a dangerous element.

[0067] In addition, the display unit (300) can display the entire work site by simultaneously displaying video images collected through multiple imaging units (100) in a screen-splitting manner.

[0068] In the case of the display unit (300), it is provided as a magnetic type monitor and can be attached to the driver's seat of the vehicle as shown in FIGS. 1 and 2, but it can also be attached to a desired location at the driver's discretion.

[0069] Additionally, the display unit (300) can receive visual information via wireless communication to receive data from the imaging unit (100).

[0070] In the case of the imaging unit (100), alarm unit (200) and display unit (300), they may be of a wireless type using magnets, making installation and retrieval easy.

[0071] In addition, in the case of a signalman alarm unit (500) corresponding to a signalman receiving alarm, it can be linked with an alarm unit (200) and an imaging unit (100) via wireless communication (e.g., Bluetooth communication, etc.) to provide an alarm to the signalman.

[0072] The heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention further includes a storage unit (400).

[0073] In the case of the storage unit (400), it receives information about a specific site captured through the imaging unit (100) and stores information about a specific site.

[0074] The storage unit (400) stores captured video and also stores driving video, and can store data corresponding to about 10 days' worth of time information.

[0075] Additionally, the storage unit (400) is conceptually distinct and can actually be formed inside the display unit (300).

[0076] The storage unit (400) may support internal or external storage and may also support expandable options.

[0077] In the case of the storage unit (400), storage and processing of data may be possible to effectively manage a large amount of data of the received shooting information.

[0078] The heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention further includes a signalman warning unit (500).

[0079] In the case of the signalman alarm unit (500), it warns the signalman of the approach of a vehicle. Since the signalman controlling traffic and personnel near heavy equipment is most exposed to danger, the present invention ensures the safety of the signalman by providing an alarm bell to the signalman.

[0080] The signalman alarm unit (500) may provide a notification to the signalman by detecting the signalman’s biometric information through the detection of biometric information based on the information captured through the imaging unit (100).

[0081] The signalman can recognize the approach of a vehicle through an alarm and take immediate action, and the driver can also reduce the vehicle's speed or stop upon receiving the warning signal.

[0082] The imaging unit (100) of the heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to the present invention is configured to include a shooting unit (110) and a bio-information detection unit (120).

[0083] In the case of the filming unit (110), the work environment corresponding to a designated site is filmed through a camera.

[0084] In addition, the camera unit (110) may photograph the entire scene and may detect a human body at the scene through human body detection.

[0085] In the case of the bio-information detection unit (120), bio-information is detected during the work environment captured through the shooting unit (110).

[0086] As illustrated in FIG. 5, the biometric information detection unit (120) can distinguish between people and objects in a captured image through a deep learning-based image analysis algorithm.

[0087] The biometric information detection unit (120) analyzes, for example, the presence of a pedestrian and movement patterns in a captured image.

[0088] The biometric information detection unit (120) detects basic object recognition, danger detection, and the presence of pedestrians, and transmits the detected image to the display unit (300).

[0089] In addition, it can recognize multiple workers or pedestrians and analyze the location and behavior of each pedestrian individually.

[0090] Conventional alarm systems are of the proximity sensor type and have the problem of causing fatigue to drivers and surrounding workers by sounding warnings indiscriminately without distinguishing between people and objects, leading to avoidance of use. In contrast, the present invention provides the effect of increasing the reliability of the alarm by accurately recognizing only the human body through the bio-information detection unit (120) to issue an alarm.

[0091] In the case of the biometric information detection unit (120), when a pedestrian enters a danger zone, the pedestrian is tracked and observed, and tracking may continue until the pedestrian leaves the danger zone.

[0092] The imaging unit (100) of the heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to the present invention is configured to include a boundary adjustment unit (130).

[0093] In the case of the boundary adjustment unit (130), the recognition range for the preset boundary is freely adjusted.

[0094] In the case of the boundary adjustment unit (130), as shown in FIGS. 7 and 8, a boundary is set for a hazardous area within the work environment and the range is dynamically adjusted, thereby expanding or shrinking the size of the boundary according to changes in the work environment.

[0095] In addition, boundaries can be adjusted by considering pedestrian movement paths, heavy equipment movement routes, and environmental changes.

[0096] The boundary control unit (130) can separate specific areas within the work environment into safe zones and dangerous zones and adjust the range of the boundary.

[0097] For example, if there is little pedestrian movement, the boundary can be reduced to prevent unnecessary warnings, and if moving objects or equipment are present, the danger zone can be expanded to prevent pedestrian access in advance.

[0098] In the case of a boundary, it may correspond to a virtual boundary line used to define the range to be detected at a given site.

[0099] The preset boundary can be adjusted according to a predetermined field of the imaging unit (100).

[0100] Depending on the specific work site, changes to the work environment may occur, such as reducing the work area or adding a hazardous area; in this case, the size of the pre-set boundary can be automatically expanded or reduced.

[0101] Pre-set boundaries can be arbitrarily expanded or contracted by algorithms, etc.

[0102] These boundaries can be dynamically set by taking into account pedestrian movement paths, risk factors, etc.

[0103] The size and boundaries of danger zones can be automatically set, and, for example, alert levels can be adjusted using pedestrian location-based data.

[0104] As illustrated in FIGS. 7 and 8, a detection range around a vehicle or heavy equipment is set within a designated site, and by allowing the detection range to be dynamically adjusted, the safety and work efficiency of workers corresponding to pedestrians are ensured.

[0105] The display unit (300) of the heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention is configured to include a video receiving unit (310) and a display unit (320).

[0106] The video receiving unit (310) networks with the imaging unit and receives the field video transmitted from the imaging unit (100).

[0107] The image receiving unit (310) can receive data through wireless networking technology in order to be connected to the imaging unit (100).

[0108] The image receiving unit (310) receives images simultaneously from a plurality of imaging units (100).

[0109] The image receiving unit (310) can receive images from each direction around the vehicle (e.g., front, rear, side) as shown in FIG. 7.

[0110] The display unit (320) displays the field video received from the video receiving unit (310) to the driver.

[0111] The display unit (320) is intended to clearly display image data of the received visual information to the driver in real time.

[0112] The display unit (320) may simultaneously output data received from a plurality of imaging units (100) in a screen splitting manner, for example, may output front, rear, and side images separately.

[0113] The display unit (320) can highlight and display the location when a risk factor is detected, and may also provide a touchscreen or physical button to enable screen operation, etc.

[0114] In addition, AI analysis data can be provided visually to induce immediate response from the driver.

[0115] The alarm unit (200) of the heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention may include a warning notification unit (210).

[0116] The warning notification unit (210) can accurately distinguish between an object and biometric information based on information collected through biometric information, and can provide a warning by recognizing only the biometric information.

[0117] The warning notification unit (210) provides notification to the driver and pedestrian when an abnormality occurs regarding the preset safety information.

[0118] In the case of pre-set safety information, it may correspond to predefined standards and conditions to maintain safety at a designated site, and may be set according to the work environment, the safety status of workers, vehicle operating conditions, etc.

[0119] Pre-set safety information has standards, and it is possible to provide warnings by comparing the captured information with the standards in real time.

[0120] Notifications can support various methods, such as voice notifications and notifications via flashing lights.

[0121] When a danger is detected, a warning broadcast can be made in voice through a speaker provided in the alarm unit (100). For example, if a worker is not wearing a safety helmet, a warning such as "Please wear a safety helmet" can be provided in voice.

[0122] In addition, for example, when the boundary adjustment unit (130) is set to 4m, if a worker approaches within the range, a voice warning saying "It is dangerous" may be provided to the worker.

[0123] The notification unit (200) of the heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention provides a warning to pedestrians and drivers when the captured information exceeds a certain level of pre-set safety information.

[0124] The imaging unit (100) can track and observe pre-set safety information and track and observe the working environment and the condition of pedestrians through behavioral pattern recognition and risk prediction via AI.

[0125] For example, when the worker is not wearing a helmet, the imaging unit (100) tracks and observes the worker, and can continue tracking until the worker wears a helmet.

[0126] The imaging unit (100) learns the behavioral patterns of pedestrians through the accumulation of data and can identify and continue observing patterns with a high risk of accidents.

[0127] Among the image information collected through the imaging unit (100), the pedestrian's posture, movement, and behavior patterns are analyzed, and if abnormal signs are detected, the notification unit (200) can provide information about this and provide a warning.

[0128] In addition, for example, by recognizing the biometric information of a pedestrian at a predetermined site through an imaging unit (100) and observing abnormal phenomena regarding the pedestrian's preset safety information, if the pedestrian's movement deviates from the normal range, a warning of the abnormal phenomenon can be provided immediately.

[0129] The imaging unit (100) of the heavy equipment collision warning system that prevents human body detection collisions through AI image analysis according to the present invention may further include an information calculation unit (140) that calculates abnormal information based on the preset safety information regarding the shooting information.

[0130] The information output unit (140) can classify the state and behavior of pedestrians among the video data.

[0131] In the case of abnormal information, it may correspond to events such as not wearing a safety helmet, for example, and generally refers to distinguishing between states and actions that deviate from safety information.

[0132] For example, in the case of the information output unit (140), if a pedestrian (worker) is not wearing safety equipment, the information regarding this is distinguished and determined to be abnormal information, thereby detecting unsafe behavior of the worker.

[0133] Additionally, the information output unit (140) may provide a warning through the notification unit (200) when the shooting information exceeds a certain level of the preset safety information.

[0134] The imaging unit (100) of the heavy equipment collision warning system that prevents human body detection collisions through AI video analysis according to the present invention detects the surrounding environment of the vehicle through a plurality of shooting units (110) that photograph the work site, and combines the shooting sites of the plurality of shooting units (110) to determine abnormal movements regarding the movement of a pedestrian and provides visual information from the vehicle to the display unit (300).

[0135] As shown in FIG. 9, the imaging unit (100) combines multiple cameras to eliminate blind spots around the vehicle and prevent accidents occurring in areas that the driver cannot see.

[0136] The signalman alarm unit (500) of the heavy equipment collision alarm system that prevents human body detection collisions through AI video analysis according to the present invention is configured to include an alarm device unit (501).

[0137] In the case of the alarm device unit (501), it is provided to the signalman as a wearable device to transmit a warning signal to the signalman.

[0138] The alarm device unit (501) can be worn directly by a signalman and may be provided as a belt type that can be worn around the waist, but may also be provided in a form that can be mounted on a vest, etc.

[0139] The alarm device unit (501) can notify the signalman of a dangerous situation by conveying a danger warning through a warning sound.

[0140] In addition, specific pattern warning sounds can be provided to distinguish types of danger.

[0141] In the case of the alarm device unit (501), the signalman may be able to detect danger through auditory alarms, tactile alarms, and visual alarms, and may also have a vibration module built in for vibration warnings.

[0142] In addition, LEDs or flashlights may be equipped for visual information, and the severity of the danger may be displayed differently depending on the blinking speed or color.

[0143] The alarm device unit (501) can be linked with the alarm unit (200) via a wireless connection and can also exchange danger information with the driver.

[0144] In the case of the signalman information unit (500), a warning signal provided through the alarm device unit (501) may be transmitted to the driver when necessary.

[0145] In addition, a warning may be manually transmitted to the driver through the alarm device unit (501).

[0146] The storage unit (400) may include a database section (410), a building section (420), an o-meeting section (430), and a sanding section (440) to store biometric information.

[0147] In the case of the DB section (410), the data corresponding to the biometric information is arbitrarily separated into data information consisting of 0s and 1s.

[0148] At this time, the information to be separated is determined arbitrarily, and for arbitrary values, it is separated and stored in the storage unit (400).

[0149] In the case of the building section (420), the configuration is such that data information is stacked.

[0150] As illustrated in Fig. 13, the data information consists of multiple separate files stacked together, and in the case of the data information, each has electromagnetic waves of different wavelengths depending on the size of the information contained therein.

[0151] In the case of the omitting section (430), it is configured to arbitrarily omit multiple stacked data information.

[0152] It is advisable to store any missing information separately.

[0153] The sanding unit (440) separately transmits the missing information to the storage unit (400).

[0154] It is preferable that the information transmitted to the storage unit (400) be transmitted separately from the multiple stacked data information with omissions and the omission information.

[0155] The scope of rights of the present invention is determined by the matters described in the patent claims, and the parentheses used in the patent claims are not intended for optional limitation but for clear components, and the descriptions within the parentheses should also be interpreted as essential components. Explanation of the symbols

[0157] 100: Imaging unit 110: Filming Department 120: Biometric information sensing unit 130: Boundary adjustment section 140: Information Output Unit 200: Alarm unit 210: Warning Notification Section 300: Display unit 310: Video receiver 320: Display section 400: Storage Unit 500: Signalman alarm unit

Claims

Claim 1 A heavy equipment collision warning system that prevents human body detection collisions through AI video analysis, characterized by comprising: an imaging unit attached to an arbitrary location of a vehicle used at a predetermined site, which photographs the predetermined site from the arbitrary location and detects visual information; an alarm unit attached to the rear of the vehicle, which transmits a warning from the rear of the vehicle; and a display unit provided to the driver's seat of the vehicle, which provides the driver with information about the predetermined site photographed by the imaging unit. Claim 2 A heavy equipment collision warning system that prevents human body detection collisions through AI image analysis, characterized in that, in claim 1, the system further includes a storage unit that receives information of a predetermined site captured through the imaging unit and stores information of the predetermined site. Claim 3 A heavy equipment collision warning system that prevents human body detection collisions through AI video analysis, characterized in that, in paragraph 2, the system further includes a signalman warning unit that warns a signalman of the approach of the vehicle. Claim 4 A heavy equipment collision warning system that prevents human body detection collisions through AI image analysis, characterized in that, in paragraph 3, the imaging unit comprises: a capturing unit that captures a predetermined scene from the arbitrary location of the vehicle and captures the captured information; and a biometric information detection unit that detects biometric information from the capturing information of the predetermined scene. Claim 5 A heavy equipment collision warning system that prevents human body detection collisions through AI image analysis, characterized in that, in claim 4, the imaging unit includes a boundary adjustment unit that freely adjusts the recognition range for a preset boundary. Claim 6 A heavy equipment collision warning system that prevents human body detection collisions through AI image analysis, wherein, in claim 5, the boundary adjustment unit expands or contracts the boundary size according to changes in the work environment. Claim 7 A heavy equipment collision warning system that prevents human body detection collisions through AI video analysis, characterized in that, in claim 5, the display unit includes: a video receiving unit that networks with the imaging unit and receives the field video transmitted from the imaging unit; and a display unit that displays the field video received from the video receiving unit to the driver. Claim 8 A heavy equipment collision warning system that prevents human body detection collisions through AI video analysis, characterized in that, in claim 6, the alarm unit includes a warning notification unit that provides notification to the driver, pedestrian, and signalman when an abnormality in preset safety information occurs. Claim 9 A heavy equipment collision warning system that prevents human body detection collisions through AI video analysis, characterized in that, in claim 7, the signalman warning unit includes a warning device portion provided to the signalman as a wearable device to transmit a warning signal to the signalman. Claim 10 A heavy equipment collision warning system that prevents human body detection collisions through AI video analysis, characterized in that, in claim 9, the imaging unit detects the surrounding environment of the vehicle through a plurality of imaging units that photograph the work site, combines the shooting sites of the plurality of imaging units to determine abnormal movement regarding the movement of the pedestrian, and provides visual information from the vehicle to the display unit.

Images