KR-102961443-B1 - FIRE DETECTION SYSTEM USING ARTIFICIAL INTELLIGENCE

Abstract

The present invention relates to a fire detection system using artificial intelligence, and more specifically, comprises: a shooting module including a thermal imaging camera and a visible light camera; a learning judgment unit that receives shooting data from the visible light camera and calculates the emissivity of an object captured by the visible light camera through image analysis of the shooting data; and a correction unit that calculates a correction value of the temperature of the object measured by the thermal imaging camera using the emissivity.

Inventors

• 정종우
• 우성주
• 이재현

Assignees

• 시티아이랩 주식회사

Dates

Publication Date

20260507

Application Date

20250314

Claims (5)

1. A shooting module including a thermal imaging camera and a visible light camera; A learning judgment unit that receives shooting data from the above visible light camera and calculates the emissivity of an object captured by the above visible light camera through image analysis of the above shooting data; and It includes a correction unit that obtains a correction value for the temperature of the object measured by the thermal imaging camera using the above emissivity, and The above learning judgment unit analyzes a training image using an artificial intelligence algorithm to match and store material names and material information, receives an emissivity matching a material name from a client, matches it with the material name, and stores it. The above learning judgment unit obtains material information from the image of the above-mentioned captured data using an artificial intelligence algorithm, transmits the emissivity that matches the material information, and A precursor symptom detection unit that receives thermal image data from the thermal image camera and, if the temperature change over time in a large area including the object or a small area including a specific part of the object in the thermal image data is greater than a preset reference change, processes the large area or the small area as a bounding box to generate first predicted data; and The system further includes a flame and smoke detection unit that learns flame images and fire smoke images, receives the captured data from the visible light camera, and generates second predicted data by processing the partial area where flames or smoke are generated in the captured data as a bounding box. It further includes a calculation unit that receives the first predicted data or the second predicted data and receives fire occurrence data or fire non-occurrence data from a client within a preset time. The above calculation unit is, Receive fire occurrence data from the client within a preset time while not receiving the first predicted data or the second predicted data, and If fire occurrence data is not received from the client within a preset time while the above-mentioned first predicted data or second predicted data is not received, fire non-occurrence data is generated, and Calculate the harmonic mean by classifying into cases where fire occurrence data is received while the second predicted data is generated (True Positive), where fire non-occurrence data is received while the second predicted data is generated (False Negative), where fire occurrence data is received while the second predicted data is not generated (False Positive), and where fire non-occurrence data is received while the second predicted data is not generated (True Negative). A fire detection system characterized in that the flame and smoke detection unit, when generating the second predicted data and receiving the fire non-occurrence data from the calculation unit, additionally learns the captured data at the time when the second predicted data was generated as a fire non-occurrence case, and when receiving the fire occurrence data from the calculation unit without generating the second predicted data, additionally learns the captured data at the time when the fire occurrence data was generated as a fire occurrence case.
2. In claim 1, It further includes a sensing module that generates sensing data by measuring one or more of the ambient temperature, humidity, and the distance between the imaging module and the object, and The correction unit receives the detection data from the detection module, and calculates the correction value of the target body temperature using the detection data and the emissivity. An alarm unit that receives a correction value from the correction unit and generates and transmits alarm data when the correction value exceeds a preset reference value; and A fire detection system characterized by further including a mapping unit that generates mapping data by mapping the thermal image data to the shooting data when receiving the alarm data from the alarm unit.
3. delete
4. A shooting module including a thermal imaging camera and a visible light camera; A learning judgment unit that receives shooting data from the above visible light camera and calculates the emissivity of an object captured by the above visible light camera through image analysis of the above shooting data; and It includes a correction unit that obtains a correction value for the temperature of the object measured by the thermal imaging camera using the above emissivity, and It further includes an installation means comprising a first installation groove formed from the upper side to the lower side of a prefabricated speed bump installed on a road, and a second installation groove formed from the upper side to the lower side of a mounting part installed on a parking surface. The above-mentioned imaging module is, The above-mentioned visible light camera and the above-mentioned thermal imaging camera are installed, and further includes a housing that is inserted and installed in the first installation groove or the second installation groove. A fire detection system characterized by further including an access authorization unit that receives a correction value from the correction unit and, if the correction value is less than or equal to a reference value, generates an open signal and transmits it to a parking lot barrier.
5. In claim 4, It further includes an output module that receives and outputs one or more of the correction value, the shooting data, and the thermal image data. The apparatus further comprises a fastening means including a vertical portion connected to the side of the housing, a horizontal portion extending horizontally from the bottom of the vertical portion and disposed on the bottom of the first installation groove or the second installation groove, a bolt hole formed in the horizontal portion, and a bolt member penetrating the bolt hole. The above-mentioned prefabricated speed bump further includes a fitting groove formed from the upper side downward and surrounding the first installation groove when viewed in a planar view, a transparent plate formed of a transparent material that covers the first installation groove, and a fitting projection provided protruding downward from the lower side of the periphery of the transparent plate so as to be inserted and fitted into the fitting groove. The support member further comprises a tubular body fixed on the bottom of the first installation groove, a first screw thread provided on the inner surface of the tubular body, a rod body whose lower part is inserted into the hollow of the tubular body, a second screw thread provided on the outer surface of the rod body and screw-fastened with the first screw thread, and an elastic member connected to the upper part of the rod body and whose upper part is supported by the lower part of the transparent plate. The above support member further includes a groove formed from the upper side to the lower side of the rod body, and A fire detection system characterized by the above elastic member comprising a body formed of an elastic material and disposed on the upper part of the rod, and a protrusion inserted into and fitted into the groove at the lower part of the body.

Description

Fire Detection System Using Artificial Intelligence The present invention relates to a fire detection system using artificial intelligence. Recently, significant economic losses have been occurring due to electric vehicle fires in underground parking lots and other areas. Although various fire detectors exist, the false alarm rate is 99.75% due to aging and minor malfunctions, and it is estimated that 21.4 billion won in firefighting resources is wasted annually. Disaster accidents such as the Coupang logistics center fire are occurring due to frequent fire detector malfunctions, leading to a response where people do not evacuate even when the alarm sounds and turn off the alarm. The prior art, Registered Patent No. 10-2689580 (hereinafter referred to as the prior art), relates to an electric vehicle fire prevention system, and more specifically, comprises a thermal imaging camera for capturing a thermal image within a fire detection target area; a fire extinguishing agent storage container for storing a fire extinguishing agent and delivering the agent to at least one of one or more discharge units configured to discharge the agent into the target area or one or more spray units configured to spray the agent onto the underside of the vehicle; and a control unit for determining whether a fire has occurred within the target area based on the thermal image, and in response to the determination that a fire has occurred within the target area, causing the storage container to deliver the fire extinguishing agent to at least one of one or more discharge units or one or more spray units. However, thermal imaging cameras are sensitive to the surrounding environment, such as emissivity, making accurate temperature analysis difficult and presenting problems such as difficulty in accurately detecting fires depending on the surrounding conditions. FIG. 1 is a schematic block diagram illustrating a fire detection system according to one embodiment of the present invention. Figure 2 is a diagram illustrating the process of a correction unit obtaining a correction value. Figure 3 is a diagram illustrating the process of obtaining a correction value through detection data and emissivity. Figure 4 is a diagram illustrating the mapping of thermal image data and captured data. Figure 5 is a diagram illustrating the output of the first expected data. Figure 6 is a diagram illustrating the output of the second expected data. FIG. 7 is a drawing for explaining a camera module installed in the first installation groove of a prefabricated speed bump. FIG. 8 is a drawing for explaining a shooting module installed in the second installation groove of a mounting part. FIG. 9 is a drawing for illustrating additional embodiments of the present invention. The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the present invention is defined only by the scope of the claims. Throughout the entire specification, when a part is described as being "connected" to another part, this includes not only cases where they are "directly connected (including electrical connections)," but also cases where they are "indirectly connected (including electrical connections)" with other elements interposed between them. FIG. 1 is a schematic block diagram illustrating a fire detection system according to one embodiment of the present invention. A fire detection system (S) (a fire detection system using artificial intelligence) according to one embodiment of the present invention is a system for more precisely detecting the occurrence of a fire or the possibility of a fire occurring in a filming target such as an electric vehicle placed on a road or parking lot, and for convenience of explanation, it will be referred to as the ‘this system’ below. Referring to FIG. 1, the present system (fire detection system (S)) may include a shooting module (1), a learning judgment unit, and a correction unit. The shooting module (1) may include a thermal imaging camera (11) and a visible light camera (12). As will be explained in more detail in the description below, the shooting module (1) may include a housing (13) on which a thermal imaging camera (11) and a visible light camera (12) are installed. The above visible light camera (12) captures a target object (or target area) to generate capture data, and the above thermal imaging camera (11) can capture the target object (or target area) to generate thermal image data. The visible light camera (12) and the thermal imaging camera (11) are known products, so