Oil spill detection system and oil spill detection method

Abstract

This system suppresses false positives or missed detections when detecting oil leaks in areas where water has accumulated. [Solution] The oil leak detection system 1 comprises one or more cameras 11 that photograph the water surface W with infrared light, a water surface area database 20 in which the areas of the water surface W at each water level are registered, and one or more computers 2 that extract a water surface area which is the area of ​​the water surface W from at least one image acquired by the camera 11 based on the information registered in the water surface area database 20, and detect oil from the water surface area.

Description

Technical Field

[0001] Embodiments of the present invention relate to oil leakage detection technology.

Background Art

[0002] In power plants or factories, rotating machinery such as turbines or generators uses lubricating oil or machinery operated by hydraulic pressure. There is also wastewater used for machine cooling or power generation, but if leaked oil mixes with this and the wastewater is discharged outside as it is, it becomes an environmental problem, so reliable oil leakage detection is required.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] Conventionally, in power plants or factories, an oil-water separator is provided in a drainage pit, and a detector provided with an electrode measures the capacitance to detect oil. However, due to dirt or aging deterioration, false detections occur. Therefore, automatic oil detection using a non-contact sensor is required. Thus, research on a method using a camera sensor has been underway. However, since the water level in the drainage pit changes depending on the operating state of the plant and the water surface position seen from the image changes, correction is required. There is also a technique for detecting oil by capturing the fluorescence of oil using ultraviolet light, but it is necessary to keep the irradiation intensity of ultraviolet light constant and it depends on the irradiation state of ultraviolet light. Therefore, when an area other than the water surface is reflected, the possibility of false detection or detection omission cannot be eliminated.

[0005] Embodiments of the present invention have been made in consideration of these circumstances, and aim to suppress false detections or failures to detect when detecting oil leakage into an area where water is accumulated. [Means for solving the problem]

[0006] An oil leak detection system according to an embodiment of the present invention comprises one or more cameras that photograph the water surface with infrared light, a water surface area database in which the areas of the water surface at each water level are registered, and one or more computers that extract a water surface area which is the area of ​​the water surface from at least one image acquired by the camera based on the information registered in the water surface area database, and detect oil from the water surface area. [Effects of the Invention]

[0007] According to the embodiments of the present invention, when detecting an oil leak into an area where water is collected, false detections or missed detections can be suppressed. [Brief explanation of the drawing]

[0008] [Figure 1] A diagram showing the configuration of the oil leak detection system according to the first embodiment. [Figure 2] Block diagram of the oil leak detection system according to the first embodiment. [Figure 3] A functional block diagram showing the processing flow of the first embodiment. [Figure 4] A screen diagram showing an example of an input video for the first embodiment. [Figure 5] A screen diagram showing an example of an image extracted from the water surface region of the first embodiment. [Figure 6] A screen diagram showing the display mode of the display of the first embodiment. [Figure 7] A diagram showing the configuration of the oil leak detection system according to the second embodiment. [Figure 8] A functional block diagram showing the processing flow of the second embodiment. [Figure 9] A functional block diagram showing the processing flow of the third embodiment. [Figure 10] Configuration diagram showing the oil leakage detection system of the fourth embodiment. [Figure 11] Functional block diagram showing the process flow of the fifth embodiment. [Figure 12] Functional block diagram showing the process flow in the oil leakage detection system of the sixth embodiment.

Embodiments for Carrying Out the Invention

[0009] (First Embodiment) Hereinafter, embodiments of an oil leakage detection system and an oil leakage detection method will be described in detail with reference to the drawings. Note that the scales of the illustrated components may be appropriately changed to aid understanding. First, the first embodiment will be described with reference to FIGS. 1 to 6.

[0010] Reference numeral 1 in FIG. 1 represents the oil leakage detection system of the first embodiment. An oil leakage detection method is implemented using this oil leakage detection system 1.

[0011] The oil leakage detection system 1 detects whether oil is mixed in the water stored in the water storage tank 10. The water storage tank 10 is provided in a facility such as a power plant or a factory, and is a drainage pit (container) for temporarily storing drainage. Oil leaked from the equipment provided in the facility may be mixed in this water storage tank 10. The oil leakage detection system 1 is a system for detecting this leaked oil and notifying (warning) the user (administrator).

[0012] A camera 11, a water level gauge 12, and a thermometer 13 are provided in the water storage tank 10. These devices are connected to the detection computer 2.

[0013] The camera 11 photographs the water surface W of the water stored in the water storage tank 10 with infrared rays. That is, this camera 11 is an infrared camera.

[0014] In addition, the camera 11 is fixed to a predetermined structure. For example, the camera 11 is fixed to a structure without fluctuations, such as the housing 15 of the water storage tank 10, by a fixing device 14. Furthermore, the camera 11 is facing the direction of the water surface W of the water storage tank 10. And the camera 11 acquires video data with infrared rays and transmits it to the detection computer 2. Note that the video includes at least one of a still image or a moving image.

[0015] Note that a plurality of cameras 11 may be provided. Since oil floats on water, by the camera 11 photographing the water surface W, it is possible to detect whether oil is mixed in the water stored in the water storage tank 10.

[0016] Since the intensity of infrared rays is different between water and oil, water and oil can be discriminated by acquiring video data with infrared rays. This infrared ray is generated because the emissivity is different between water and oil, so lighting may or may not be provided. Note that the intensity includes at least one of the reflection intensity or the radiation intensity.

[0017] The water level gauge 12 is provided inside the water storage tank 10. The water level gauge 12 measures the water level, which is the height of the water surface W of the water stored in the water storage tank 10. Note that a plurality of water level gauges 12 may be provided.

[0018] The water level gauge 12 acquires the height of the water surface W of the water storage tank 10 by either a contact type or a non-contact type method and transmits it to the detection computer 2.

[0019] The thermometer 13 is provided near the water storage tank 10. The thermometer 13 measures at least one of the temperature at the location where the water surface W is present or the ambient air temperature. Note that a plurality of thermometers 13 may be provided.

[0020] The thermometer 13 measures the water temperature of the water storage tank 10 or the ambient air temperature thereof and transmits the data to the detection computer 2. At this time, when only the air temperature can be acquired, the current water temperature is estimated from the air temperature. The water temperature can be estimated, for example, from the design data of the water storage tank 10 based on the air temperature.

[0021] As shown in Figure 2, the oil leak detection system 1 includes a camera 11, a water level gauge 12, a thermometer 13, and a detection computer 2.

[0022] The detection computer 2 comprises an input unit 3, an output unit 4, a communication unit 5, a processing circuit 6, and a storage unit 7.

[0023] The oil leak detection system 1 consists of a computer equipped with hardware resources such as a CPU (Central Processing Unit), GPU (Graphics Processing Unit), ROM (Read Only Memory), RAM (Random Access Memory), HDD (Hard Disk Drive), and SSD (Solid State Drive). The CPU executes various programs, enabling software-based information processing using these hardware resources. Furthermore, the oil leak detection method is implemented by having the computer execute various programs.

[0024] Each component of the oil spill detection system 1 does not necessarily have to be installed on a single computer. For example, one oil spill detection system 1 may be implemented using multiple computers connected to each other via a network. For instance, various databases may be installed on separate computers.

[0025] Furthermore, the configuration of the oil leak detection system 1 may be implemented as a cloud service. In other words, the computers that make up the oil leak detection system 1 may be servers on the cloud. For example, not only the configuration that performs memory processing, but the entire configuration that performs the main processing may reside on the cloud, and the user may only configure the oil leak detection system 1 and check its input / output via an API (Application Programming Interface) or a web browser.

[0026] The input unit 3 receives predetermined information in response to the user's operation of the detection computer 2. This input unit 3 includes input devices such as a mouse, keyboard, and touch panel. In other words, predetermined information is input to the detection computer 2 in response to the operation of these input devices.

[0027] The output unit 4 includes a display 8 (Figure 3) for displaying information. In other words, the detection computer 2 includes a display 8 (display device) for outputting analysis results and displaying images. The display 8 may be separate from the computer body or integrated into it.

[0028] Although a display 8 (Figure 3) for displaying images is shown as an example of the output unit 4, other configurations are also possible. For example, an image may be displayed using a head-mounted display or a projector. Furthermore, a printer for printing information on paper may be used instead of the display 8. In other words, the output unit 4 includes a head-mounted display, a projector, or a printer.

[0029] The communication unit 5 communicates with other computers via a communication line such as the Internet. While the detection computer 2 and the other computers are connected via the Internet, other configurations are also possible. For example, the detection computer 2 and the other computers may be connected via a LAN (Local Area Network), WAN (Wide Area Network), or mobile communication network. Alternatively, each device may be connected to the others via a bus.

[0030] The processing circuit 6 is, for example, a circuit equipped with a CPU, GPU, or a dedicated or general-purpose processor. This processor realizes various functions by executing various programs stored in the memory unit 7. The processing circuit 6 may also be composed of hardware such as an FPGA (Field Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit). Various functions can also be realized by this hardware. Furthermore, the processing circuit 6 can realize various functions by combining software processing by the processor and programs with hardware processing.

[0031] The processing circuit 6 has the function of executing various processes (Figure 3). These functions are realized by the CPU executing programs stored in memory or the HDD.

[0032] The memory unit 7 stores a predetermined program to be executed by the processing circuit 6. The memory unit 7 also stores various information necessary for oil detection. Furthermore, the memory unit 7 stores various information necessary for oil detection based on information stored in various databases.

[0033] As shown in Figure 3, the oil leak detection system 1 includes a water surface area database 20, a structure database 21, a temperature-based video database 22, and a detection result database 23. These various databases are collections of information organized so that they can be stored in memory, HDD, or cloud computing resources, and can be searched or accumulated. For example, these various databases are stored in the storage unit 7 (Figure 2).

[0034] The water surface area database 20 has pre-registered information indicating the area of ​​the water surface W at each water level in the water storage tank 10. Based on this information, the water surface area, which is the area of ​​the water surface W, is extracted from the video acquired by the camera 11.

[0035] For example, the water surface area database 20 has pre-registered information indicating which areas within the camera's field of view 11 contain the water surface W for each water level. From this water surface area database 20, data indicating the area of ​​the water surface W can be obtained according to the water level.

[0036] In this case, it is not necessary for all water levels to be registered; the water level may be determined by interpolation using linear interpolation or similar methods from values ​​before and after the required water level. Alternatively, with the fixed position of camera 11 and the field of view of camera 11 pre-set, the position of the water surface W projected into the field of view of camera 11 may be determined from the 3D model of the water tank 10 and the water level information.

[0037] The structure database 21 contains pre-registered 3D information on structures located within the camera's field of view. These structures include those that make up the water tank 10 and those that do not make up the water tank 10.

[0038] For example, the structure database 21 has pre-registered distance information indicating where a structure is projected within the camera 11's field of view. From this structure database 21, data indicating the region of structures above the water level can be obtained. Alternatively, with the fixed position of the camera 11 and the camera 11's field of view information pre-set, structures above the water surface W that are projected within the camera 11's field of view can be determined from the 3D model of the structure and the water level information.

[0039] The temperature-based video database 22 has pre-registered values ​​for each pixel in the video footage of the water surface W captured by the camera 11, corresponding to each pixel at different temperatures for water and oil, which are the objects to be identified.

[0040] For example, the temperature-based video database 22 contains data of water captured using infrared light for each water temperature. The temperature-based video database 22 contains multiple videos of water captured using infrared light for each water temperature. From these videos, the pixel values ​​for each temperature can be obtained. In addition to water, the temperature-based video database 22 may also contain data of oil used in the plant captured using infrared light for each water temperature.

[0041] The detection computer 2 obtains the current water temperature from the thermometer 13, retrieves the pixel value of the infrared image corresponding to the current water temperature from the temperature-specific image database 22, identifies this value as a threshold indicating the water surface W, and further identifies any area different from this threshold as the oil surface Q.

[0042] Next, the process flow for the oil leak detection system 1 of the first embodiment to detect oil in the water storage tank 10 will be described.

[0043] First, infrared images of the water surface W captured by camera 11, the water level measured by water level gauge 12, information registered in the water surface area database 20, and information registered in the structure database 21 are obtained.

[0044] The detection computer 2 performs preprocessing 30. This preprocessing 30 includes water surface area extraction processing 31 and structure exclusion processing 32.

[0045] In the water surface region extraction process 31, the detection computer 2 extracts a water surface region from at least one video image acquired by the camera 11, based on the water level measured by the water level gauge 12 and the information registered in the water surface region database 20.

[0046] The detection computer 2 may, without using the water level measured by the water level gauge 12, extract the water surface region, which is the area of ​​the water surface W, from at least one image acquired by the camera 11, based solely on the information registered in the water surface region database 20.

[0047] In the structure exclusion process 32, the detection computer 2 excludes the area of ​​the structure from the area of ​​the water surface W in the video based on the information registered in the structure database 21.

[0048] The water surface region extraction process 31 and the structure exclusion process 32 are processes that are executed in relation to each other. For example, in the water surface region extraction process 31, the detection computer 2 uses the video data from the camera 11 and the water level information from the water level gauge 12 to obtain the water surface region from the water surface region database 20 and extract the region of the water surface W from the video data. Furthermore, in the structure exclusion process 32, the detection computer 2 excludes the region of structures obtained from the structure database 21 and obtains video in which only the water surface W is captured.

[0049] For example, as shown in Figure 4, there is an image acquired by camera 11. This image shows the water surface W and the oil surface Q floating on the water surface W. Because the infrared intensity differs between water and oil, the regions of the water surface W and the oil surface Q have different pixel values. For example, the water surface W is shown in dark gray, and the oil surface Q is shown in light gray. In addition, structures other than the water surface W, such as the main body 15, are also reflected in this image. Note that the structures include not only the main body 15 but also predetermined objects located near or inside the water storage tank 10 that are reflected in the image. Furthermore, the region of the water surface W in the image changes according to the water level.

[0050] Here, as shown in Figure 5, the detection computer 2 specifies an exclusion region E to leave the water surface W region while excluding the structure region. This exclusion region E is generated by the water surface region extraction process 31 and the structure exclusion process 32. The detection computer 2 then excludes the exclusion region E from the video, that is, it masks the exclusion region E, and generates a video in which only the water surface W region is visible.

[0051] As shown in Figure 3, the detection computer 2 performs threshold discrimination processing 33. At this point, the temperature measured by the thermometer 13 is input to the detection computer 2 (Figure 2).

[0052] In the threshold discrimination process 33, the detection computer 2 estimates the water temperature of the water surface W from the temperature measured by the thermometer 13. Furthermore, the detection computer 2 sets a threshold for identifying oil from the values ​​of each pixel registered in the temperature-specific video database 22, corresponding to the estimated water temperature. Then, based on this threshold, the detection computer 2 identifies the oil surface Q from the region of the water surface W in the video. In other words, the detection computer 2 detects oil from the region of the water surface W.

[0053] Next, the detection computer 2 registers the detection result indicating the presence or absence of oil in the water surface area in the detection result database 23. Here, the detection result database 23 registers registration information including the detection result indicating the presence or absence of oil in the water surface area, the image of the water surface area, and the temperature measured by the thermometer 13, associated with the time the registration information was acquired. Note that the time includes the date.

[0054] Next, the detection computer 2 displays on the display 8, in a manner that allows comparison between the registered information associated with any one time registered in the detection result database 23 and the registered information associated with other times. The detection computer 2 receives multiple times to be displayed from the user and retrieves the registered information associated with these times from the detection result database 23.

[0055] In addition, the detection result database 23 may be registered with the video data, the temperature measured by the thermometer 13, and the pixel values ​​for each temperature obtained from the temperature-specific video database 22, all of which are associated with each other.

[0056] For example, as shown in Figure 6, the detection computer 2 displays images from different times side by side on the display 8. These images include the time, the temperature measured by the thermometer 13, and the detection result. If the detection result indicates the presence of oil, a predetermined notification (warning) may be given to the user.

[0057] Furthermore, the display 8 may show a warning when oil is detected, as well as video footage of the oil detection and data obtained by infrared photography of water at the temperature or temperature used as the basis for the detection.

[0058] According to the first embodiment, the water surface W to be detected can be extracted from the image from camera 11, water level, and structural information. Then, the threshold can be automatically changed based on the temperature and type of oil to detect the oil.

[0059] In particular, by using a camera 11 that captures images using infrared light, and by combining the position information of the camera 11 with the position information of the water level gauge 12 and the structure, it is possible to extract only images of the water surface area. Furthermore, by automatically changing the threshold for extracting the oil area based on temperature information, false detections or missed detections can be suppressed.

[0060] (Second Embodiment) Next, a second embodiment will be described with reference to Figures 7 to 8. Note that components identical to those shown in the previously described embodiment are denoted by the same reference numerals, and redundant descriptions are omitted.

[0061] As shown in Figure 7, the camera 11 of the second embodiment photographs the water surface W of the water stored in the water tank 10 using infrared and visible light. Alternatively, the camera 11 may be a visible light camera that photographs the water surface W using only visible light.

[0062] The oil leak detection system 1 of the second embodiment includes, in addition to the configuration of the first embodiment described above, an illumination device 16 that emits infrared and visible light illumination.

[0063] Multiple lighting devices 16 may be provided. Furthermore, the lighting device 16 includes an infrared lighting device that emits infrared illumination light and a visible light lighting device that emits visible light illumination light. In addition, the lighting device 16 may consist of either an infrared lighting device or a visible light lighting device.

[0064] Furthermore, the lighting fixture 16 is fixed to a predetermined structure. For example, the lighting fixture 16 is fixed to a stable structure such as the frame 15 of the water storage tank 10 using a fixing device 17. In addition, the lighting fixture 16 is oriented toward the water surface W of the water storage tank 10.

[0065] Next, the process flow for the oil leak detection system 1 of the second embodiment to detect oil in the water storage tank 10 will be described.

[0066] As shown in Figure 8, the oil leak detection system 1 of the second embodiment performs a floating object detection process 34 and an area exclusion process 35 in addition to the processing of the first embodiment described above.

[0067] First, infrared and visible light images of the water surface W captured by camera 11, the water level measured by water level gauge 12, information registered in the water surface area database 20, and information registered in the structure database 21 are obtained. Then, the detection computer 2 performs preprocessing 30.

[0068] Next, the detection computer 2 performs a floating object detection process 34. Here, the detection computer 2 extracts areas of floating objects other than water and oil from the image of the water surface W captured in visible light. Floating objects are, for example, objects such as algae or debris.

[0069] Next, the detection computer 2 performs area exclusion processing 35. Here, the detection computer 2 excludes areas of floating matter from the area of ​​the water surface W extracted from the image of the water surface W captured with infrared light. Then, the detection computer 2 performs threshold discrimination processing 33 and detects oil from the area of ​​the water surface W.

[0070] The detection computer 2 may acquire information on the color range of structures or algae from the visible light image, identify the structures or algae, and extract the area of ​​the water surface W that should be detected.

[0071] Furthermore, the detection computer 2 may compare the visible light image with a predetermined template image to identify structures or algae and extract the area of ​​the water surface W. In addition, the detection computer 2 may use machine learning-based object detection or segmentation techniques to identify structures or algae and extract the area of ​​the water surface W.

[0072] According to the second embodiment, even if the value of each pixel in the image changes due to temperature or type of oil, the threshold can be changed, thereby enabling proper detection of oil. Furthermore, even if existing structures or algae appear within the image's field of view and have similar values ​​to oil, false detection or detection failure can be suppressed.

[0073] Furthermore, even when infrared images alone provide insufficient information, visible light images can be used to identify the water surface W. Additionally, oil detection becomes possible without the need to pre-input information about interfering objects. When oil is detected, the dimensional information from both infrared and visible light images may be expanded and processed accordingly. In this case, various databases would also undergo similar information expansion.

[0074] Furthermore, the presence of the lighting device 16 allows for the use of reflectance information in addition to infrared emissivity, making the difference between water and oil clearer and thus easier to distinguish.

[0075] (Third embodiment) Next, a third embodiment will be described with reference to Figure 9. Note that components identical to those shown in the previously described embodiments are denoted by the same reference numerals, and redundant descriptions are omitted.

[0076] The oil leak detection system 1 of the third embodiment includes, in addition to the configuration of the first embodiment described above, a captured video database 24.

[0077] The video database 24 contains multiple videos of the water surface W captured by camera 11. For example, the video database 24 contains multiple videos in chronological order.

[0078] Next, the process flow for the oil leak detection system 1 of the third embodiment to detect oil in the water storage tank 10 will be described.

[0079] The oil leak detection system 1 of the third embodiment performs a best image extraction process 36 in addition to the processing of the first embodiment described above.

[0080] First, infrared images of the water surface W captured by the infrared and visible light camera 11 are stored in the captured image database 24. Next, the water level measured by the water level gauge 12, the information registered in the water surface area database 20, the information registered in the structure database 21, and the information registered in the captured image database 24 are obtained. Then, the detection computer 2 executes the best image extraction process 36.

[0081] In the best image extraction process 36, the detection computer 2 extracts the best image from multiple images registered in the captured image database 24, the image with the smallest fluctuation component of the water surface W.

[0082] Next, the detection computer 2 performs preprocessing 30. In this preprocessing 30, the detection computer 2 extracts the water surface W region from the best image based on the information registered in the water surface region database 20. Then, the detection computer 2 performs threshold discrimination processing 33 to detect oil from the water surface W region.

[0083] If water is constantly flowing into the water surface W, its surface is agitated, and this agitation may cause false detection or detection failure. However, according to the third embodiment, by extracting the best image and then extracting the region of the water surface W from this best image, false detection or detection failure can be suppressed.

[0084] Alternatively, the detection computer 2 may reduce the effects of shaking by accumulating video footage at regular intervals, such as one second, in the captured video database 24 and generating an average video from that data. Alternatively, the video with the smallest difference from the average video over a certain period may be extracted.

[0085] Furthermore, when detecting the movement of the water surface W in the video, and when detecting the component with the least fluctuation from the video, the detection computer 2 may use the information from the water level gauge 12.

[0086] Alternatively, an IMU (Inertial Measurement Unit) may be floated on the water surface W. The detection computer 2 may then detect the fluctuation component of the water surface W from the acceleration obtained from the IMU.

[0087] (Fourth Embodiment) Next, a fourth embodiment will be described with reference to Figure 10. Note that components identical to those shown in the previously described embodiments are denoted by the same reference numerals, and redundant descriptions are omitted.

[0088] The oil leak detection system 1 of the fourth embodiment includes, in addition to the configuration of the first embodiment described above, a float 18 that floats on the water surface W of the water storage tank 10 and whose height changes in conjunction with the water level.

[0089] The camera 11 is fixed to the float 18 with a fixing device 14, and its relative position and orientation to the water surface W are fixed. Therefore, the height of the camera 11 moves in conjunction with the water level. In other words, the positional relationship between the water surface W and the camera 11 remains constant.

[0090] Furthermore, if multiple cameras 11 are provided, multiple floats 18 may be provided. In addition, one camera 11 may be supported by multiple floats 18.

[0091] According to the fourth embodiment, the positional relationship between the water surface W and the camera 11 remains constant, which enables high-precision extraction of the water surface W region.

[0092] In addition, since the positional relationship between the camera 11 and the structure may change, or algae may appear in the image captured by the camera 11, pre-processing 30 (Figure 3) may be performed in the fourth embodiment.

[0093] (Fifth embodiment) Next, a fifth embodiment will be described with reference to Figure 11. Note that components identical to those shown in the previously described embodiments are denoted by the same reference numerals, and redundant descriptions are omitted.

[0094] The oil leak detection system 1 of the fifth embodiment comprises a trained model 40 and a reference video database 25.

[0095] The trained model 40 has been pre-trained to detect oil from regions of the water surface W extracted from images of the water surface W taken with infrared light.

[0096] The trained model 40 comprises an input layer, a hidden layer, and an output layer. Input data is input to the input layer. The hidden layer's parameters are pre-trained using training data. The output layer outputs output data that shows the results of processing in the hidden layer in response to the input data input to the input layer.

[0097] The trained model 40 is machine-learned using training data in which at least one of the input data or data simulating it is input, and at least one of the output data or data simulating it is output.

[0098] The reference video database 25 has reference video footage of the oil-free water surface W region that was filmed in the past and is pre-registered.

[0099] Next, the process flow for the oil leak detection system 1 of the fifth embodiment to detect oil in the water storage tank 10 will be described.

[0100] The oil leak detection system 1 of the fifth embodiment performs an oil detection process 37 instead of the threshold determination process 33 (Figure 3) of the first embodiment described above.

[0101] In the oil detection process 37, the detection computer 2 calculates the difference between a reference image registered in the reference image database 25 and the image of the target to be detected captured by the camera 11, and based on this difference, detects oil in the area of ​​the water surface W in the image from the camera 11. For example, the detection computer 2 obtains a reference image of the water surface W from the reference image database 25 when the water temperature (temperature) and water level are the same as when the water surface W was photographed by the camera 11. Then, the detection computer 2 calculates the difference from this reference image and detects oil.

[0102] Furthermore, in the oil detection process 37, the detection computer 2 inputs the image of the water surface W region as input data to the trained model 40 and outputs a detection result indicating the presence or absence of oil as output data. The input data may also include the temperature measured by the thermometer 13 (Figure 3).

[0103] If sufficient video footage has been accumulated from camera 11 installed in the same location, normal video footage taken in the past when no oil was detected will be registered in the reference video database 25 along with water temperature and water level information.

[0104] The detection computer 2 then inputs the video from camera 11, along with the water temperature and water level, into the trained model 40. For example, a normal model trained using machine learning, such as an autoencoder, detects areas of normal water surface W, and areas of abnormal water surface W are detected as areas of oil surface Q.

[0105] Alternatively, infrared images of water and oil taken in advance, along with the water temperature at that time, may be acquired. Information indicating that the water surface W is correct (training label) may be added to these acquired images, and a trained model 40 (Figure 11) that distinguishes between water and oil may be generated in advance using machine learning. Then, oil may be detected by this trained model 40.

[0106] Furthermore, in the oil detection process 37, the detection computer 2 obtains video footage from the reference video database 25 under the same conditions as the current video, including water temperature and water level, and detects oil based on the difference between this reference video and the current video. In this case, if the camera 11 is fixed to the float 18 (Figure 10), the detection computer 2 may first determine the position of the camera 11 with high accuracy using SLAM (Simultaneous Localization and Mapping), and then perform the process of calculating the difference.

[0107] According to the fifth embodiment, since oil can be detected from the water surface W region using machine learning technology, detection accuracy can be further improved.

[0108] In the fifth embodiment, the oil detection process 37 performs both the process of detecting oil by calculating the difference from the reference image and the process of detecting oil using the trained model 40, but either one of these processes may be performed.

[0109] The trained model 40 may also be one whose parameters have been pre-trained to detect oil from regions of the water surface W extracted from images of the water surface W taken in visible light.

[0110] (Sixth Embodiment) Next, the sixth embodiment will be described with reference to Figure 12. Note that components identical to those shown in the previously described embodiments are denoted by the same reference numerals, and redundant descriptions are omitted.

[0111] In the sixth embodiment of the oil leak detection system 1 shown in Figure 12, in the pre-processing 30 of the process flow for detecting oil in the water storage tank 10, the detection computer 2 performs image output range setting processing 41 and brightness difference correction processing 42 in addition to water surface area extraction processing 31 and structure exclusion processing 32.

[0112] In the water surface region extraction process 31 of the preprocessing 30, the detection computer 2 extracts water surface regions that are areas of the water surface W from the images acquired by the camera 11 that photographs the water surface W with infrared light, based on the information registered in the water surface region database 20 (information on the area of ​​the water surface W at each water level).

[0113] Next, in the image output range setting process 41 of the preprocessing 30, the detection computer 2 sets the image output range of the camera 11 to a fixed range based on the brightness of a reference object placed within the shooting range of the camera 11, and such that the brightness difference between water and oil in the extracted water surface region becomes clear.

[0114] In other words, while camera 11 (especially the long-wavelength infrared camera) is less affected by visible light, if ambient light such as reflections from high-temperature objects or direct sunlight is reflected on the water surface, localized brightness anomalies may occur, potentially causing the reflected direct sunlight on the water surface W to be mistaken for oil. Furthermore, at night when illumination conditions are low, the ambient temperature decreases, reducing the temperature difference between water and oil, which may result in a smaller brightness difference between water and oil on the water surface W.

[0115] In contrast, by setting the image output range of the camera 11 to the fixed range described above, the difference in brightness between water and oil in the water surface region becomes clear, making it possible to reduce the influence of ambient light and illumination conditions. Here, the reference object is a target (object) that has a known brightness or emissivity that does not deviate from the brightness or emissivity range of water and oil, is placed within the shooting range of the camera 11, and is not excluded by the structure exclusion process 32 of the preprocessing 30.

[0116] Next, the detection computer 2 performs the brightness difference correction process 42 of the preprocessing 30 in combination with the image output range setting process 41 described above. In other words, in this brightness difference correction process 42, the detection computer 2 performs a correction process in real time to increase the brightness difference between water and oil in the water surface region after structures have been removed by the structure exclusion process 32. This brightness difference correction process 42 includes, for example, histogram extension, local contrast enhancement, gamma correction, and radiation correction, and is particularly effective when the brightness difference between water and oil in the water surface region is small, making it possible to improve the separation of water and oil.

[0117] Furthermore, in preprocessing 30, if there are high-brightness areas within the shooting range of the camera 11 that are significantly brighter than other areas (for example, heated walls or metal surfaces in the summer), the detection computer 2 removes or corrects the high-brightness areas in the water surface database 20, as there is a risk that the gradation of the water surface area may be compressed by the fixed range setting by the image output range setting process 41, and updates the water surface database 20.

[0118] Then, in the threshold discrimination process 43, the detection computer 2 uses a luminance threshold representing the water surface W to distinguish between water and oil in the water surface region and detects oil from the water surface region.

[0119] According to the sixth embodiment, the image output range of the camera 11 is set to a fixed range based on the brightness of a reference object placed within the shooting range of the camera 11, and such that the brightness difference between water and oil in the water surface region becomes clear. Therefore, even when ambient light such as reflections from high-temperature objects or direct sunlight is reflected on the water surface, or when lighting conditions are low, such as at night, the brightness difference between water and oil in the water surface region extracted from the image of the camera 11 becomes clear, thus reducing the influence of ambient light and lighting conditions, and suppressing false detections and missed detections when detecting oil on the water surface W.

[0120] Furthermore, by performing a brightness difference correction process 42 on the water surface region after the structure exclusion process 32, which increases the brightness difference between water and oil, the separation of water and oil can be improved even when the brightness difference between water and oil is minute. As a result, when combined with the image output range setting process 41, false detections and missed detections of oil on the water surface W can be further suppressed.

[0121] Furthermore, if a high-brightness area exists within the shooting range of the camera 11, the removal or correction of this high-brightness area is performed on the water surface area database 20, thereby suppressing a decrease in the accuracy of oil detection.

[0122] Furthermore, the detection computer 2 may automatically adjust the fixed range setting value performed in the image output range setting process 41 in real time based on the brightness of the reference object, and by referring to the water level measured by the water level gauge 12 and the temperature measured by the thermometer 13 as needed. This improves the accuracy of oil detection even in the face of environmental changes such as temperature changes due to day and night or seasons, and fluctuations in water level.

[0123] Although the present invention has been described above based on the first to sixth embodiments, a configuration applied in any one embodiment may be applied to another embodiment, or the configurations applied in each embodiment may be combined.

[0124] In the embodiments described above, the determination of an arbitrary value (pixel value) using a reference value (threshold) may also be a determination of "whether the arbitrary value is greater than or equal to the reference value." Alternatively, this determination may also be a determination of "whether the arbitrary value exceeds the reference value." Alternatively, this determination may also be a determination of "whether the arbitrary value is less than or equal to the reference value." Alternatively, this determination may also be a determination of "whether the arbitrary value is less than or equal to the reference value." Furthermore, the reference value may not be fixed but may change. Therefore, instead of a reference value, a predetermined range of values ​​may be used, and the determination may be made as to whether the arbitrary value falls within the predetermined range. Alternatively, the errors that occur in the device may be analyzed in advance, and a predetermined range including the error range centered on the reference value may be used for determination.

[0125] Note that the oil leak detection system 1 may include components other than those shown in the block diagram, and some of the components shown in the block diagram may be omitted.

[0126] Note that the arrows in the functional block diagram are just one example of a processing flow, and there may be other processing flows besides those indicated by the arrows. Also, the order of each process is not necessarily fixed, and the order of some processes may be reversed. Furthermore, some processes may be executed in parallel with other processes.

[0127] The aforementioned oil leak detection system 1 comprises a control device, a storage device, an output device, an input device, and a communication interface. Here, the control device includes a highly integrated processor such as a CPU (Central Processing Unit), GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), or dedicated chip. The storage device includes ROM (Read Only Memory), RAM (Random Access Memory), HDD (Hard Disk Drive), SSD (Solid State Drive), etc. The output device includes a display panel, head-mounted display, projector, printer, etc. The input device includes a mouse, keyboard, touch panel, etc. This oil leak detection system 1 can be implemented with a hardware configuration using a standard computer.

[0128] The program or learned model 40 executed by the aforementioned oil leak detection system 1 is provided pre-installed in ROM or the like. Additionally or alternatively, this program or learned model 40 is provided as an installable or executable file stored on a computer-readable non-temporary storage medium. This storage medium includes CD-ROMs, CD-Rs, memory cards, DVDs, flexible disks (FDs), and the like.

[0129] Furthermore, the program or trained model 40 executed by this oil spill detection system 1 may be stored on a computer connected to a network such as the Internet and provided for download via the network. In other words, the program or trained model 40 may be provided from cloud computing resources. Alternatively, a server on the cloud may execute the program or trained model 40, and only the processing results may be provided via the cloud. In addition, this system can also be configured by connecting and combining separate modules, each independently performing the function of its components, via a network or dedicated line.

[0130] In the example described above, the computer constituting the oil leak detection system 1 is shown to perform various processes (including various processes such as discrimination, judgment, evaluation, estimation, and setting), but other configurations are also possible. For example, the user may perform some of the aforementioned processes, and the computer may receive the input of the processing results and use them in its own processing.

[0131] According to at least one embodiment described above, the oil leak detection system 1 includes one or more cameras 11 that capture images of the water surface W using infrared light, thereby suppressing false detections or missed detections when detecting an oil leak into an area where water is collected.

[0132] While several embodiments of the present invention have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These embodiments can be carried out in various other forms, and various omissions, substitutions, modifications, and combinations are possible without departing from the spirit of the invention. These embodiments or their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. A singular expression is not necessarily intended to limit to just one thing, and a singular expression may refer to multiple things. [Explanation of Symbols]

[0133] 1…Oil spill detection system, 2…Detection computer, 3…Input unit, 4…Output unit, 5…Communication unit, 6…Processing circuit, 7…Storage unit, 8…Display, 10…Water tank, 11…Camera, 12…Water level gauge, 13…Thermometer, 14…Equipment, 15…Structure, 16…Lighting equipment, 17…Equipment, 18…Float, 20…Water surface area database, 21…Structure database, 22…Temperature-specific video database, 23…Detection result data 24...Database of captured video, 25...Reference video database, 30...Preprocessing, 31...Water surface region extraction process, 32...Structure exclusion process, 33...Threshold discrimination process, 34...Floating object detection process, 35...Region exclusion process, 36...Best image extraction process, 37...Oil detection process, 40...Trained model, 41...Image output range setting process, 42...Brightness difference correction process, 43...Threshold discrimination process, E...Excluded region, Q...Oil surface, W...Water surface.

Claims

1. One or more cameras that photograph the water surface using infrared light, A water surface area database in which the water surface area at each water level is registered, Based on the information registered in the water surface area database, one or more computers extract a water surface area from at least one image acquired by the camera and detect oil from the water surface area, Equipped with, Oil spill detection system.

2. The camera has a structure database in which three-dimensional information of structures located within its field of view is registered. The aforementioned computer, Based on the information registered in the aforementioned structure database, the area of ​​the structure is excluded from the water surface area, and oil is detected from the water surface area. It is structured in such a way. The oil leak detection system according to claim 1.

3. The system includes one or more water level gauges for measuring the water level, which is the height of the water surface. The aforementioned computer, Based on the water level measured by the water level gauge and the information registered in the water surface area database, the water surface area is extracted from the camera image and oil is detected from the water surface area. It is structured in such a way. An oil leak detection system according to claim 1 or claim 2.

4. One or more thermometers for measuring at least one of the temperature of the water surface or the ambient temperature, A detection result database is registered in association with the time the registration information was acquired, and includes a detection result indicating the presence or absence of oil in the water surface area, an image of the water surface area, and the temperature measured by the thermometer. A display that shows information, Equipped with, The computer is configured to display on the display, in a manner that allows comparison between the registered information associated with any one of the time periods registered in the detection result database and the registered information associated with other time periods. An oil leak detection system according to claim 1 or claim 2.

5. One or more thermometers for measuring at least one of the temperature of the water surface or the ambient temperature, A temperature-based video database in which the values ​​of each pixel at different temperatures for water and oil, corresponding to each pixel of the video footage of the water surface captured by the aforementioned camera, are registered. Equipped with, The aforementioned computer, The water surface temperature is estimated from the temperature measured by the thermometer, Based on the estimated water temperature, a threshold for identifying oil is set from the value of each pixel registered in the temperature-specific video database. Based on the threshold, oil is detected from the water surface region. It is structured in such a way. An oil leak detection system according to claim 1 or claim 2.

6. The camera includes a camera that photographs the water surface in visible light, The aforementioned computer, From the image of the water surface captured with visible light, the region of floating matter other than water and oil floating on the water surface is extracted. From the image of the water surface captured by the infrared light, the area of ​​floating matter is excluded from the water surface region, and oil is detected from the water surface region. It is structured in such a way. An oil leak detection system according to claim 1 or claim 2.

7. The system includes a video database in which multiple videos of the water surface captured by the aforementioned camera are registered. The aforementioned computer, From the multiple videos registered in the aforementioned video database, the best video with the least amount of water surface fluctuation component is extracted. Based on the information registered in the aforementioned water surface area database, the water surface area is extracted from the best image. To detect oil from the water surface region, It is structured in such a way. An oil leak detection system according to claim 1 or claim 2.

8. Equipped with one or more floats whose position changes in conjunction with the water level, The camera includes a camera that is fixed to the float and whose position and orientation relative to the water surface are fixed. An oil leak detection system according to claim 1 or claim 2.

9. The system includes a pre-trained model whose parameters have been pre-machine-trained to detect oil from the water surface region extracted from the image of the water surface captured by the infrared light, The aforementioned computer, The trained model is given the image of the water surface region as input data and outputs the detection result indicating the presence or absence of oil as output data. It is structured in such a way. An oil leak detection system according to claim 1 or claim 2.

10. It includes a reference image database in which reference images of the aforementioned water surface area in an oil-free state, which were taken in the past, are registered. The aforementioned computer, The difference between the reference video and the video of the object to be detected captured by the camera is calculated. Based on the aforementioned difference, oil is detected from the water surface region. It is structured in such a way. An oil leak detection system according to claim 1 or claim 2.

11. The system includes one or more lighting devices that emit the aforementioned infrared illumination light. An oil leak detection system according to claim 1 or claim 2.

12. One or more cameras that photograph the water surface using infrared light, A water surface area database in which the water surface area at each water level is registered, Based on the information registered in the aforementioned water surface area database, the water surface area is extracted from at least one image acquired by the camera, The video output range of the aforementioned camera is set to a fixed range based on the brightness of a reference object placed within the camera's shooting range, and such that the brightness difference between water and oil in the water surface region becomes clear. One or more computers that detect oil from the water surface region, Equipped with, Oil spill detection system.

13. The aforementioned computer, In combination with setting the video output range of the aforementioned camera to a fixed range, By performing a brightness difference correction process on the water surface region to amplify the brightness difference between water and oil, oil can be detected from the water surface region. It is structured in such a way. The oil leak detection system according to claim 12.

14. The aforementioned computer, If a high-luminance region exists within the camera's shooting range that is significantly brighter than other regions, this high-luminance region is removed or corrected in the water surface region database. It is structured in such a way. The oil leak detection system according to claim 12 or claim 13.

15. One or more cameras that photograph the water surface using infrared light, A water surface area database in which the water surface area at each water level is registered, One or more computers, This method uses The computer extracts a water surface region from at least one image acquired by the camera, based on information registered in the water surface region database, and detects oil from the water surface region. Oil spill detection method.

16. One or more cameras that photograph the water surface using infrared light, A water surface area database in which the water surface area at each water level is registered, One or more computers, This method uses The computer extracts a water surface region, which is the water surface region, from at least one image acquired by the camera, based on the information registered in the water surface region database. The video output range of the aforementioned camera is set to a fixed range based on the brightness of a reference object placed within the camera's shooting range, and such that the brightness difference between water and oil in the water surface region becomes clear. To detect oil from the water surface region, Oil spill detection method.

Images