JP2013181795A - Floating object detection device and floating object detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a floating object detection device and floating object detection method capable of properly detecting a floating object.SOLUTION: The floating object detection device includes a light reception part for imaging a water surface, an area extraction part 48 for extracting an area coincident with an extraction condition of a floating object in a water surface image imaged by the light reception part, a storage part 45 for storing a template image for detecting the floating object, a similarity calculation part 44 for calculating a similarity between an upper area hardly affected by a state of waves around the floating object of the extracted area and the stored template image, and a floating object detection part 46 for detecting the area extracted by the area extraction part 48 as an area with the floating object existing therein when the calculated similarity is higher than a predetermined value.

Description

  The present invention relates to a suspended matter detection apparatus and a suspended matter detection method.

2. Description of the Related Art Conventionally, a monitoring device using a laser radar is known as a monitoring device that is installed in a ship or the like and detects floating matters on the water surface (see, for example, Patent Document 1).
A monitoring device using such a laser radar irradiates the periphery of the ship with laser light from the laser radar, the laser light reaches the water surface and is reflected, and the reflected light is imaged by the light receiving unit, The captured image signal is displayed on the display unit.

JP 2009-244212 A

As a method for detecting floating objects on the water surface based on an image obtained by imaging the reflected light of pulsed laser light, for example, the shape of the floating object is stored in advance as a template image, and the captured image and the template image are stored. A method is conceivable in which similarity is calculated by comparison, and suspended matter is detected based on the calculated similarity.
However, the shape of the floating object captured as an image changes depending on the state of waves around the floating object. Therefore, even if the shape of the floating object is stored in advance as a template image, there is a problem that the floating object cannot be detected properly depending on the state of the waves around the floating object.

  The present invention has been made in view of such circumstances, and even when the shape of a floating object captured as an image changes depending on the state of waves around the floating object, the floating object is appropriately removed. It is an object of the present invention to provide a floating substance detection device and a floating substance detection method capable of detection.

In order to achieve the above object, the present invention employs the following means.
The floating substance detection apparatus according to the present invention is a floating substance detection apparatus that detects a floating substance floating on water, the imaging means for imaging the water surface, and the floating substance from the image of the water surface imaged by the imaging means. Extraction means for extracting an area that matches the object extraction condition, storage means for storing a template image for detecting the floating substance, and conditions of waves around the floating substance in the area extracted by the extraction means A similarity calculation unit that calculates a similarity between the upper region that is not easily affected by the template image stored in the storage unit, and a similarity calculated by the similarity calculation unit is higher than a predetermined value Detecting means for detecting an area extracted by the extracting means as an area where the suspended matter exists.

  According to the floating object detection apparatus according to the present invention, an area that matches the extraction condition of the floating substance is extracted from the captured image of the water surface, and the influence of the state of waves around the floating substance in the extracted area is affected. The degree of similarity between the difficult-to-receive upper area and the template image is calculated, and the extracted area is detected as the area where the suspended matter exists when the calculated degree of similarity is higher than a predetermined value. By doing this, the upper region excluding the lower region that is easily affected by the state of waves around the floating material from the region extracted as the region that matches the extraction condition of the floating material is defined as the template image. The similarity can be calculated. Therefore, even when the shape of the floating object picked up as an image changes depending on the state of waves around the floating object, the floating object can be detected appropriately.

  In the suspended matter detection apparatus according to the first aspect of the present invention, the storage means stores a plurality of template images respectively corresponding to a plurality of wave heights, and the floating matter detection apparatus has input means for inputting the wave height of the water surface. A selection means for selecting a template image corresponding to the wave height input by the input means from a plurality of template images stored in the storage means, and the similarity calculation means is extracted by the extraction means The degree of similarity between the upper region of the region and the template image selected by the selection unit is calculated.

  The floating object detection apparatus according to the first aspect of the present invention stores a plurality of template images corresponding to a plurality of wave heights, selects a template image corresponding to the input wave height, and selects the selected template image. Is the target of similarity calculation. By doing so, it is possible to detect an appropriate suspended matter according to the wave height.

  In the suspended matter detection apparatus according to the second aspect of the present invention, the similarity calculation unit includes the upper region above the center of gravity of the region extracted by the extraction unit, and the template image stored in the storage unit. The degree of similarity is calculated. By doing this, the region above the center of gravity of the extracted region is the target for calculating the similarity with the template image, and the shape of the floating object that is captured as an image is determined by the state of waves around the floating object. Even if it changes, suspended matter can be detected appropriately.

  The floating substance detection apparatus according to the third aspect of the present invention is configured to determine a relative position of the region detected by the detection unit from the imaging unit based on a position in the image of the water surface of the region extracted by the extraction unit. And a position specifying means for specifying. By doing in this way, the position of the area detected by the detection means can be specified as a relative position from the imaging means.

  The floating substance detection apparatus according to the fourth aspect of the present invention includes a recognition means for recognizing the position of the floating substance detection apparatus, and the water surface of the area extracted by the extraction means with the position of the area detected by the detection means. And a position specifying means for specifying the position based on the position of the floating object detecting device recognized by the recognition means. By doing in this way, the position of the area detected by the detecting means can be specified as an absolute position.

  The floating substance detection apparatus according to the fifth aspect of the present invention includes an area specifying unit that specifies an area having a luminance higher than a predetermined luminance from the water surface image, and an area for calculating an area of the area specified by the area specifying unit. A calculating means, wherein the extracting means extracts a region in which the area calculated by the area calculating means matches a predetermined area condition from the regions specified by the region specifying means. To do. By doing in this way, the area | region which becomes an area which agree | coincides with a predetermined area condition can be extracted as a candidate of the area | region where a floating substance exists.

  In the above-described floating object detection apparatus according to the fifth aspect, the floating object detection apparatus includes a peripheral length calculation unit that calculates a peripheral length of a region extracted as a region that matches the area condition, and the extraction unit includes a region that matches the area condition. A region in which the perimeter calculated by the perimeter calculation unit matches a predetermined perimeter condition may be extracted from the extracted regions. By doing so, it is possible to extract a region that matches a predetermined area condition and has a peripheral length that matches a predetermined peripheral length condition as a candidate for a region in which a suspended matter exists.

  A floating object detection apparatus according to a sixth aspect of the present invention includes a laser light source that emits laser light, and the imaging unit picks up the reflected light of the laser light source as an image, and the laser light emitted by the laser light source. The wavelength is 1400 nm or more and 2600 nm or less. By using laser light having such a wavelength, reflection of the laser light from the water surface can be suppressed, and reflected light from a suspended matter floating on the water can be imaged more clearly. Therefore, it is possible to improve the detection accuracy of floating substances floating on the water.

  The floating substance detection method according to the present invention is a floating substance detection method for detecting a floating substance floating on the water, the imaging step of imaging the water surface, and the floating surface from the image of the water surface imaged by the imaging step An extraction step for extracting a region that matches the extraction condition of the object, an upper region that is less affected by the state of waves around the floating substance in the region extracted by the extraction step, and for detecting the floating substance A similarity calculation step for calculating a similarity with a template image, and a region where the floating substance is present in the region extracted by the extraction step when the similarity calculated by the similarity calculation step is higher than a predetermined value And a detection step of detecting as a feature.

  According to the floating substance detection method according to the present invention, an area that matches the extraction condition of the floating substance is extracted from the captured image of the water surface, and the influence of the state of waves around the floating substance in the extracted area is affected. The degree of similarity between the difficult-to-receive upper area and the template image is calculated, and the extracted area is detected as the area where the suspended matter exists when the calculated degree of similarity is higher than a predetermined value. By doing this, the upper region excluding the lower region that is easily affected by the state of waves around the floating material from the region extracted as the region that matches the extraction condition of the floating material is defined as the template image. The similarity can be calculated. Therefore, even when the shape of the floating object picked up as an image changes depending on the state of waves around the floating object, the floating object can be detected appropriately.

  According to the present invention, even when the shape of a floating object captured as an image changes depending on the state of waves around the floating object, the floating object detection device and the floating object can be detected appropriately. An object detection method can be provided.

It is a block diagram which shows the whole structure of the floating body detection apparatus of 1st Embodiment. It is the top view which looked at the suspended | floating matter detection apparatus of 1st Embodiment from the water surface upper direction. It is the side view which looked at the suspended | floating matter detection apparatus of 1st Embodiment from the side. It is a block diagram which shows the structure of the image processing apparatus of 1st Embodiment. It is a flowchart which shows the process which the image processing apparatus of 1st Embodiment performs. It is a figure which shows an example of the image imaged by the imaging process of FIG. It is a figure which shows an example of the image averaged by the averaging process of FIG. It is a figure which shows an example of the image binarized by the binarization process of FIG. It is a figure which shows an example of the area | region extracted by the area | region extraction process of step S506 of FIG. It is a figure which shows an example of the area | region extracted by the area | region extraction process of step S508 of FIG. It is a figure which shows an example of the floating body detected by the floating body detection process of FIG. It is a figure which shows an example of the template image memorize | stored in a memory | storage part. It is a figure which shows the area | region (upper area | region) used as the calculation object of similarity. It is a figure which shows the image imaged by the light-receiving part, and a target.

[First Embodiment]
Below, the floating substance detection apparatus of 1st Embodiment of this invention is demonstrated with reference to drawings.
FIG. 1 is a block diagram showing the overall configuration of the suspended matter detection apparatus according to the first embodiment of the present invention.
As shown in FIG. 1, the floating object detection device includes a laser radar 1, a laser radar control unit 2, a control device 3, and a display device 4. The laser radar 1 includes a light transmitting unit 11 and a light receiving unit 12.

The light transmitting unit 11 includes a semiconductor laser 111 that emits laser light, a light transmitting lens 112 that irradiates the target with the laser light emitted from the semiconductor laser 111, and light transmission for adjusting the angle of the light transmitting lens 112. A lens actuator (not shown) is provided.
The semiconductor laser 111 is a laser having a wavelength of 1400 nm or more and 2600 nm or less (for example, 1550 nm) having a high water absorption, and is supplied with power from a laser power supply 26 in the laser radar control unit 2 to be described later, Eject. The light transmission lens actuator adjusts the position of the light transmission lens 112 based on a control signal supplied from a laser radar control unit 2 described later. As a result, the angle of the laser light incident on the light transmission lens 112 can be adjusted, and the laser light can be irradiated in a desired range. Note that the wavelength of the semiconductor laser 111 is set to the above-described range because, when a laser having a wavelength of less than 1400 nm or more than 2600 nm is used, there is a wavelength band in which the absorption of laser light into water is low and reflected in water. This is because it becomes difficult to detect the suspended matter with respect to the water surface around the suspended matter that is the imaging target.

  The light receiving unit 12 includes a zoom lens 121 and an ICCD (image intensifier CCD) camera head 123. The zoom lens 121 collects the reflected light that is irradiated from the light transmitting unit 11 and reflected by the water surface and the floating matter floating on the water surface. The ICCD camera head 123 converts the light collected by the zoom lens 121 into an electric signal to generate an image signal, and outputs the image signal to the image processing device 25 in the laser radar control unit 2. In the first embodiment, the light receiving unit 12 and the image processing device 25 constitute an imaging device (imaging means). Such a laser radar 1 has a structure in which the rotation angle and the attack angle are adjusted to desired angles by the swivel base 5.

  The laser radar control unit 2 controls the light transmitting unit 11 of the laser radar 1, the light receiving unit 12, and the turntable 5 that adjusts the rotation angle of the laser radar 1 based on various control signals supplied from the control device 3. To do. The laser radar control unit 2 includes a turntable drive unit 21, a synchronization circuit 22, a control signal conversion device 23, an image processing device 25, and a laser power source 26.

  The control device 3 generates various control signals for controlling the laser radar 1, outputs the generated various control signals to the laser radar control unit 2, and displays an image supplied from the laser radar control unit 2 on the display device 4. Output to. Specifically, the control device 3 controls the laser radar control unit 2 so that laser light is emitted from the semiconductor laser 111. The control device 3 outputs a synchronization control signal necessary for emitting laser light to the laser radar control unit 2.

  The synchronization control signal output from the control device 3 is supplied to the synchronization circuit 22 via the control signal conversion device 23 in the laser radar control unit 2. The synchronization circuit 22 generates a synchronization signal for synchronizing laser light transmission and light reception based on the input synchronization control signal, and outputs this synchronization signal to the laser power source 26. The laser power source 26 generates an operation signal of the semiconductor laser 111 in the light transmission unit 11 included in the laser radar 1 based on the synchronization signal supplied from the synchronization circuit 22, and drives the semiconductor laser 111 based on the operation signal. To do.

  When the semiconductor laser 111 is driven by the laser power source 26, laser light is emitted from the semiconductor laser 111. The laser light is expanded to an irradiation area having a predetermined range by the light transmitting lens 112 and emitted to the outside, and the laser light reflected by the water surface to be the laser light irradiation area and the floating matter floating on the water surface is received by the light receiving unit. Guided to 121. Then, the reflected light guided to the light receiving unit 12 is imaged by the ICCD camera head 123. The reflected light imaged by the ICCD camera head 123 is output as an image signal to the image processing device 25 in the laser radar control unit 2.

  Next, the arrangement of the imaging device (the light receiving unit 12 and the image processing device 25) of the floating object detection device of the first embodiment and the area of the water surface imaged by the imaging device will be described with reference to FIGS. . FIG. 2 is a plan view of the floating substance detection device viewed from above the water surface, and FIG. 3 is a side view of the floating substance detection device viewed from the side.

As shown in FIGS. 2 and 3, the floating object detection apparatus of the first embodiment has a configuration in which a laser radar 1 is mounted on a ship body 100. The laser radar 1 has a horizontal viewing angle of θ ₁ and a vertical viewing angle of θ ₂ and images the imaging region 30 as an image signal. When the position of the laser radar 1 is L0, the position of L2 is the position closest to the laser radar 1 in the imaging range, and the position of L1 is the position farthest from the laser radar 1 in the imaging range. In the image signal output to the image processing device 25, the L2 side is the lower region of the image, and the L1 side is the upper region of the image.

  As shown in FIG. 3, the laser radar 1 is installed in the ship main body 100 so that the laser beam irradiation position of the laser radar 1 is at a height H from the water surface. The axis from the laser beam irradiation position of the laser radar 1 to the center of the imaging region is φ from the horizontal line passing through the irradiation position. This angle φ becomes the depression angle (camera depression angle) of the light receiving unit 12 including the ICCD camera head 123.

Also, when the target object in the imaging region 30 imaged by the light receiving unit 12 a (floating, etc.) is T, the azimuth angle of the target T in the plan view of FIG. 2 is a gamma _t. The angle from the horizontal line of the target T (depression angle of the target) is theta _t in side view in FIG.

Next, processing executed by the image processing unit 25 of the first embodiment will be described with reference to FIGS.
FIG. 4 is a block diagram illustrating a configuration of the image processing device 25. FIG. 5 is a flowchart illustrating processing executed by the image processing apparatus 25 according to the first embodiment. FIG. 6 is a diagram illustrating an example of an image captured by the imaging process of FIG. FIG. 7 is a diagram illustrating an example of an image averaged by the averaging process of FIG. FIG. 8 is a diagram illustrating an example of an image binarized by the binarization process of FIG. FIG. 9 is a diagram illustrating an example of a region extracted by the region extraction processing in step S506 of FIG. FIG. 10 is a diagram illustrating an example of a region extracted by the region extraction process in step S508 of FIG. FIG. 11 is a diagram illustrating an example of the suspended matter detected by the suspended matter detection process of FIG. 5. FIG. 12 is a diagram illustrating an example of a template image stored in the storage unit. FIG. 13 is a diagram illustrating a region (upper region) that is a target of similarity calculation.

  FIG. 4 is a block diagram illustrating a configuration of the image processing device 25. The image processing apparatus 25 according to the first embodiment includes an averaging filter 40, a binarization filter (region specifying unit) 41, an area calculating unit (area calculating unit) 42, and a perimeter calculating unit (peripheral length calculating unit). 43, a similarity calculation unit (similarity calculation unit) 44, a storage unit (storage unit) 45, a suspended matter detection unit (detection unit) 46, a wave height input unit (input unit) 47, and a region extraction unit ( An extraction unit) 48, a template selection unit (selection unit) 49, and a position specifying unit (position specifying unit) 50. Each unit excluding the storage unit 45, the wave height input unit 47, and the template selection unit 49 functions as an image processing unit that processes an input image. Each unit other than the storage unit 45, the wave height input unit 47, and the template selection unit 49 may be configured as a plurality of hardware modules that execute the respective processes, or may be executed by one hardware module such as a CPU. It may be configured as a plurality of software modules. Details of processing executed by each unit will be described later.

FIG. 5 is a flowchart showing processing executed by the image processing device 25.
In step S501, the image processing device 25 determines whether or not the wave height is input by the operator of the floating object detection device at the wave height input unit 47. If it is determined that the wave height is input, the process proceeds to step S502. The wave height input unit 47 inputs, as a wave height, a numerical value input by an operator of the floating object detection apparatus via an input device such as a keyboard. The operator of the floating object detection device confirms the state of the water surface with the naked eye and inputs the wave height.

In step S <b> 502, the image processing device 25 executes imaging processing by acquiring reflected light captured by the ICCD camera head 123 as an image signal. An example of an image captured by the imaging process in step S502 is an image as shown in FIG.
In step S503, the averaging filter 40 averages the image acquired in step S502, and generates an averaged image. An example of the image averaged by the averaging process in step S503 is an image as shown in FIG. In the image of FIG. 7, a minute size white area (spike noise) that was present in the image of FIG. 6 is removed.

  Note that the averaging filter 40 calculates the average value of the luminance around the pixel of interest (for example, eight pixels around the pixel of interest), and executes a process of replacing the calculated average value of luminance with the luminance of the pixel of interest. . The processing for the target pixel is executed for all the pixels of the image while switching the target pixel one by one. Even if there is a pixel whose luminance is greatly different from the surrounding pixels by the averaging process executed by the averaging filter 40, the luminance is averaged with the surrounding luminance, and an image with a gradual change in luminance is output.

  In step S504, the binarization filter 41 performs binarization processing on the image averaged in step S503, and generates a binarized image. An example of an image binarized by the binarization process in step S504 is an image as shown in FIG. In the image of FIG. 8, a region having a luminance higher than a predetermined luminance is specified among the pixels included in the image of FIG. In FIG. 8, the portion specified as an area higher than the predetermined luminance by the binarization process is shown in black.

  In step S505, the area calculation unit 42 calculates the area of the region extracted by the binarization process in step S504. Here, the area does not correspond to the area of each region on the image itself, but corresponds to the actual area of each region on the water surface. As shown in FIG. 3, the laser radar 1 is arranged so that the height H from the water surface is the irradiation position of the laser beam, and the camera depression angle of the light receiving unit 12 is φ. Therefore, the upper area of the image and the lower area of the image have different areas when converted to the actual area on the water surface even if the area on the image is the same. Specifically, when the area on the image is the same in the area A existing in the upper part of the image and the area B existing in the lower part of the image, the area A on the water surface is larger. Therefore, the area calculation unit 42 calculates the area on the water surface in consideration of the area of the region on the image extracted by the binarization process in step S504 and the position of each region on the image.

  In step S506, the region extraction unit 48 determines whether or not the area calculated by the area calculation processing in step S505 matches a predetermined area condition, and extracts a region that matches the predetermined area condition. Here, the predetermined area condition refers to a condition that is stored in advance in the storage unit 45 in consideration of the floating object to be detected by the floating object detection device. For example, it is assumed that a certain range of area suitable for the suspended matter to be detected is set as the area. In the example illustrated in FIG. 8, it is determined that the region 8 a and the region 8 b do not meet a predetermined area condition, and a region other than the region 8 a and the region 8 b is extracted by the region extraction unit 48. An example of the region extracted by the region extraction processing in step S506 is shown in FIG. In FIG. 9, the region 8a and the region 8b are removed from the plurality of regions extracted in FIG.

  In step S507, the perimeter calculation unit 43 calculates the perimeter of the region extracted by the region extraction process in step S506. The perimeter here refers not to the perimeter of each area on the image itself but to the actual perimeter on the water surface of each area. As shown in FIG. 3, the laser radar 1 is arranged such that the height H from the water surface is an irradiation position of the laser light, and the depression angle of the light receiving unit 12 is φ. Therefore, the upper area of the image and the lower area of the image have different perimeters in terms of the actual perimeter on the water surface even if the perimeter on the image is the same. Specifically, when the peripheral length on the image is the same in the region A existing in the upper part of the image and the region B existing in the lower part of the image, the peripheral length on the water surface is longer in the region A. Therefore, the perimeter calculation unit 43 calculates the perimeter on the water surface in consideration of the perimeter of the region on the image extracted by the region extraction processing in step S506 and the position of each region on the image. .

  In step S508, the region extraction unit 48 determines whether or not the perimeter calculated by the perimeter calculation processing in step S507 matches a predetermined perimeter condition, and the region meets the predetermined perimeter condition. To extract. Here, the predetermined peripheral length condition refers to a condition that is stored in advance in the storage unit 45 in consideration of the floating object to be detected by the floating substance detection device. For example, a certain range of perimeters suitable for the suspended object to be detected is set as the perimeter. In the example shown in FIG. 9, it is determined that the region 9a, the region 9b, and the region 9c do not meet the predetermined perimeter condition, and the region extraction unit 48 extracts other regions other than the region 9a, the region 9b, and the region 9c. Is done. An example of the region extracted by the region extraction processing in step S508 is shown in FIG. In FIG. 10, the region 9a, the region 9b, and the region 9c are removed from the plurality of regions extracted in FIG.

  In step S509, the template selection unit 49 selects a template image corresponding to the input wave height from a plurality of template images stored in the storage unit 45 based on the wave height input in step S501. Here, an example of the template image stored in the storage unit 45 is shown in FIG. 12A shows a template image selected when the wave height is higher than H2, FIG. 12B shows a template image selected when the wave height is higher than H1 and lower than H2, and FIG. 12C shows the template image selected. The template image selected when a wave height is below H1 is shown. When the wave height is high, the suspended matter may float significantly from the water surface, but when the wave height is low, the suspended matter will not float significantly from the water surface. Therefore, when the wave height is higher than H2, the template image (a) that can detect the suspended matter even if the wave height rises greatly is selected. On the other hand, when the wave height is equal to or lower than H1, the template image (c) that can detect the suspended matter when the wave height does not rise significantly from the water surface is selected. When the wave height is higher than H1 and lower than H2, the template image (b) is selected.

  In step S510, the similarity calculation unit 44 calculates the similarity between the region extracted in step S508 and the template image selected in step S509. In calculating the similarity, the similarity calculation unit 44 does not set the entire area extracted in step S508 as the similarity calculation target. The similarity calculation unit 44 uses the region above the region among the regions extracted in step S508 as the similarity calculation target. The reason for calculating the similarity in the upper region is that it is not easily affected by the state of waves around the floating object. In other words, it is more suitable to calculate the similarity by removing the lower region that is easily affected by the state of the waves around the suspended matter. Since the lower area is affected by the state of waves around the floating object and the image is likely to change, by excluding this lower area, the upper area is subject to similarity calculation, Even when the shape of a floating object captured as an image changes depending on the situation or the like, the similarity can be calculated in order to appropriately detect the floating object.

  FIG. 10 shows three regions, region 10a, region 10b, and region 10c, as the regions extracted in step S508. Then, the similarity calculation unit 44 specifies an area surrounded by a dotted line in FIG. 13A as the upper area of the image whose similarity is to be calculated. In addition, the similarity calculation unit 44 specifies an area surrounded by a dotted line in FIG. 13B as the upper area of the image whose similarity is to be calculated. Further, the similarity calculation unit 44 specifies an area surrounded by a dotted line in FIG. 13C as the upper area of the image whose similarity is to be calculated.

  In step S508, various methods can be used as a method for specifying the upper region from the extracted regions. For example, it is possible to use a method for specifying a region above the center of gravity of the region extracted in step S508. Further, for example, a method of specifying a certain proportion of the area in the height direction from the upper part of the area extracted in step S508 (for example, an area of 50% of the area height) can be used.

  Note that, when calculating the similarity, the similarity calculation unit 44 enlarges or reduces the size of the template image according to the position of the extracted region in step S508 that is the similarity calculation target. For example, when a template image having a size suitable for the center position of the image is stored in the storage unit 45, in step S508, if the extracted region is an upper region, the template image is reduced. Used to calculate similarity. If the extracted region is a region existing in the lower part of the image, the template image is enlarged and used for calculating the similarity. By doing in this way, it is possible to calculate an appropriate similarity according to the position on the image of the region extracted in step S508.

  Furthermore, when calculating the similarity, the similarity calculation unit 44 calculates the similarity multiple times while changing the angle of the template image, and the highest similarity is obtained among the multiple calculations. The similarity in the case is used as the output result. By doing in this way, even if it is a case where the angle of the floating object in a template image differs from the angle of the actually imaged floating object, a similarity degree can be calculated appropriately.

In step S511, the suspended matter detection unit 46 determines whether or not the similarity calculated in step S510 is higher than a predetermined value. If it is determined that the similarity is higher than the predetermined value, the similarity is higher than the predetermined value. The determined area is detected as an area where floating substances exist. If the similarity calculated in step S510 is lower than the predetermined value, it is assumed that the area where the suspended matter is present is not detected. Note that the predetermined value is stored in advance in the storage unit 45 in consideration of the suspended matter to be detected.
In FIG. 11, it is shown that the area 10a is detected as an area where floating substances are present from the three areas of FIG.

In step S512, the image processing apparatus 25 determines whether or not a region where a floating substance exists is detected in step S511. If it is determined that a region where a floating substance exists is detected, the process proceeds to step S513. .
On the other hand, when it is not determined that the region where the suspended matter exists is detected, the processing after the imaging processing in step S502 is executed again.

  In step S513, the position specifying unit 50 specifies the relative position of the floating object from the laser radar 1 based on the position on the image of the area detected as the area where the floating substance exists in step S511. Each process of the flowchart of FIG.

Next, the process in which the area calculating unit 42 calculates the area in step S505 and the process in which the position specifying unit 50 specifies the position in step S513 will be described in detail with reference to FIG. FIG. 14 is a diagram illustrating an image captured by the light receiving unit 12 and a region (hereinafter, target T) indicating the target on the image. The image picked up by the light receiving unit 12 is an image having X _max pixels in the horizontal direction and Y _max pixels in the vertical direction. In FIG. 14, the upper left is the origin of the X and Y coordinates, and the range from the origin to the coordinates (X _max , Y _max ) is an image corresponding to the imaging region 30 in FIG. 2.

The target T in FIG. 14 corresponds to the region extracted in step S504 in FIG. 5, the width in the horizontal direction (X-axis direction) is Xw pixels, and the length in the vertical direction (Y-axis direction) is also Xw. This is a pixel area. Coordinates X _t next on the X-axis of the center position of the target T, the coordinate of the Y-axis strip becomes Y _t. Then, the following conditional expression holds for the target T on the image captured by the light receiving unit 12 and the actual target.

S _pix / (X _w ) ² = S _t / (T _w ) ² (1)
_Tw / W = _Xw / _Xmax (2)
_{W = 2 · L t · tan} (θ 1/2) (3)
L _t = H / tan (θ _t ) (4)
θ _t = φ + θ ₂ · (Y _t −Y _max / 2) / Y _max (5)
γ _t = θ ₁ · (X _t −X _max / 2) / X _max (6)
Here, S _pix : Area of the target T on the image [pixel], S _t : Actual area of the target T [m ² ], X _w : Width of the target T on the image [pixel], T _w : actual width [m] of the target T, W: width [m] of the imaging region 30 at the position of the target T, L _t : distance [m] between the light receiving unit 12 and the target T, θ ₁ : light reception Viewing angle [degrees] in the horizontal direction of the section 12, θ ₂ : Viewing angle in the vertical direction of the light receiving section 12 [degree], θ _t : Depression angle [degree] of the target T, γ _t : Azimuth angle of the target [degree] ], Φ: Camera depression angle [degrees] of the laser radar 1, X _t : X coordinate on the image of the target T, Y _t : Y coordinate on the image of the target T, X _max : Number of pixels in the horizontal direction of the image [Pixel], Y _max : Number of pixels in the vertical direction of the image [Pixel], H: Height of the laser beam irradiation position of the laser radar 1 from the water surface [m].

In step S505 of FIG. 5 described above, the area calculation section 42 calculates the actual area S _t of the target T on the basis of (6) From the above conditional expression (1). In step S513 of FIG. 5 described above, the position specifying unit 50 determines the azimuth angle γ _{t of the} target and the distance L _t between the laser radar 1 and the target T based on the above conditional expressions (1) to (6). And the relative position of the target T (floating matter) from the laser radar 1 is specified.

  As described above, the floating substance detection apparatus according to the first embodiment extracts a region that matches the extraction condition of the floating substance from the captured image of the water surface, and the waves around the floating substance in the extracted area. The similarity between the upper region that is not easily affected by the situation and the template image is calculated, and when the calculated similarity is higher than a predetermined value, the extracted region is detected as the region where the suspended matter exists. By doing this, the upper region excluding the lower region that is easily affected by the state of waves around the floating material from the region extracted as the region that matches the extraction condition of the floating material is defined as the template image. The similarity can be calculated. Therefore, even when the shape of the floating object picked up as an image changes depending on the state of waves around the floating object, the floating object can be detected appropriately.

In addition, the floating object detection apparatus of the first embodiment stores a plurality of template images corresponding to each of a plurality of wave heights in the storage unit 45, selects a template image corresponding to the input wave height, and selects the template image. The obtained template image is set as a target for calculating the similarity. By doing so, it is possible to detect an appropriate suspended matter according to the wave height.
In addition, the floating object detection apparatus of the first embodiment sets the relative position of the area detected by the floating substance detection unit 46 from the laser radar 1 to the position in the image of the water surface of the area extracted by the area extraction unit 48. Identify based on. In this way, the position of the area detected by the floating substance detection unit 46 can be specified as a relative position from the laser radar 1.

In addition, the floating object detection apparatus according to the first embodiment calculates the area of the area specified by the binarization filter 41 and the binarization filter 41 that specify an area having a luminance higher than a predetermined luminance from the water surface image. An area calculation unit 42, and the region extraction unit 48 extracts a region in which the area calculated by the area calculation unit 42 meets a predetermined area condition from the regions specified by the binarization filter 41 To do. By doing in this way, the area | region which becomes an area which agree | coincides with a predetermined area condition can be extracted as a candidate of the area | region where a floating substance exists.
Furthermore, the floating substance detection apparatus according to the first embodiment includes a peripheral length calculation unit 43 that calculates the peripheral length of a region extracted as a region that matches the area condition, and the region extraction unit 48 includes a region that matches the area condition. Are extracted from the extracted regions as the peripheral length calculated by the peripheral length calculation unit 43 in accordance with a predetermined peripheral length condition. By doing so, it is possible to extract a region that matches a predetermined area condition and has a peripheral length that matches a predetermined peripheral length condition as a candidate for a region in which a suspended matter exists.

  In the floating object detection apparatus of the first embodiment, the light receiving unit 12 captures the reflected light of the laser light source as an image, and the wavelength of the laser light emitted by the laser light source is 1400 nm or more and 2600 nm or less. By using laser light having such a wavelength, reflection of the laser light from the water surface can be suppressed, and reflected light from a suspended matter floating on the water can be imaged more clearly. Therefore, it is possible to improve the detection accuracy of floating substances floating on the water.

[Second Embodiment]
The floating object detection apparatus of the first embodiment is based on the wave height input in step S501 of FIG. 5, and one template corresponding to the input wave height from among a plurality of template images stored in the storage unit 45. Select an image. In the first embodiment, the selected one template image is a target of similarity calculation.
In contrast, the second embodiment selects a plurality of template images from a plurality of template images stored in the storage unit 45 when the wave height is high.

  For example, when the wave height is higher than H2, not only the template image shown in FIG. 12 (a) but also three template images including the template images shown in FIG. 12 (b) and FIG. 12 (c) are selected. In this case, the similarity calculation unit 44 calculates the similarity between the region extracted in step S508 and the three template images. And the process of step S512 is performed using the similarity with the highest numerical value among the three similarities calculated for the three template images.

  When the wave height is higher than H2, the three template images shown in FIGS. 12A, 12B, and 12C are selected because when the wave height is higher than H2, the suspended matter is larger from the water surface. This is because the state of rising and the state of not changing change over time. Although the processing load of the image processing device 25 is increased by increasing the number of template images whose similarity is to be calculated as compared with the first embodiment, the detection accuracy of floating substances can be improved accordingly.

  If the wave height is higher than H1 and lower than H2, two template images shown in FIGS. 12B and 12C are selected in step S509. The reason why the template image shown in FIG. 12A is not selected is that the suspended object that matches the template image shown in FIG.

  If the wave height is equal to or less than H1, one template image shown in FIG. 12C is selected in step S509. The template images in FIGS. 12A and 12B are not selected because the wave height is equal to or lower than H1, and the suspended matter that matches the template images in FIGS. 12A and 12B is a laser. This is because the image is not picked up by the radar 1.

[Other Embodiments]
In the first embodiment, the relative position of the suspended object from the laser radar 1 is specified in step S513 in FIG. 5, but other modes may be used. For example, the absolute position of the suspended matter may be specified by latitude and longitude.
In this case, the floating object detection device includes a recognition device (recognition unit) capable of recognizing the latitude and longitude of its own device such as GPS (Global Positioning System).
And the floating object detection apparatus of 2nd Embodiment is the area | region detected as an area | region where a floating substance exists in step S511 of FIG. 5 of 1st Embodiment and the absolute position (latitude and longitude) recognized by the recognition apparatus. The absolute position (latitude and longitude) of the floating object detection device is specified based on the position on the image.

  In another embodiment, the floating body detection device recognizes the position of the floating body detection device, and the position of the region detected by the floating body detection unit 46 is obtained from the water surface of the region extracted by the region extraction unit 48. The identification is performed based on the position in the image and the position of the floating substance detection device recognized by the recognition device. By doing in this way, the position of the area detected by the suspended matter detection unit 46 can be specified as an absolute position (latitude and longitude).

  In 1st Embodiment, although the light-receiving part 12 imaged an image using the ICC camera head 123, another aspect may be sufficient. For example, the light receiving unit 12 may capture an image using an infrared (Infrared (IR)) camera.

DESCRIPTION OF SYMBOLS 1 Laser radar 25 Image processing apparatus 40 Average filter 41 Binarization filter 42 Area calculation part 43 Perimeter calculation part 44 Similarity calculation part 45 Storage part 46 Floating object detection part 47 Wave height input part 48 Area detection part 49 Template selection part 50 Location specification part

Claims (9)

A floating substance detection device for detecting floating substances floating on water,
Imaging means for imaging the water surface;
Extraction means for extracting a region that matches the extraction condition of the suspended matter from the image of the water surface imaged by the imaging means;
Storage means for storing a template image for detecting the suspended matter;
Similarity calculation means for calculating the similarity between the upper region of the region extracted by the extraction means that is not easily affected by the state of waves around the suspended matter and the template image stored in the storage means;
Detecting means for detecting an area extracted by the extracting means as an area where the suspended matter is present when the similarity calculated by the similarity calculating means is higher than a predetermined value;
A suspended matter detection apparatus comprising: The storage means stores a plurality of template images respectively corresponding to a plurality of wave heights;
The floating substance detection device is
Input means for inputting the wave height of the water surface;
Selecting means for selecting a template image corresponding to the wave height input by the input means from a plurality of template images stored in the storage means;
The suspended matter according to claim 1, wherein the similarity calculation unit calculates a similarity between the upper region of the region extracted by the extraction unit and the template image selected by the selection unit. Detection device.   The similarity calculation unit calculates a similarity between the upper region above the center of gravity of the region extracted by the extraction unit and the template image stored in the storage unit. The suspended | floating matter detection apparatus of Claim 1 or Claim 2. Position specifying means for specifying the relative position of the area detected by the detecting means from the imaging means based on the position in the image of the water surface of the area extracted by the extracting means;
The suspended | floating matter detection apparatus of any one of Claims 1-3 characterized by the above-mentioned. Recognizing means for recognizing the position of the floating substance detecting device;
Position specification for specifying the position of the area detected by the detection means based on the position of the area extracted by the extraction means in the image of the water surface and the position of the floating substance detection device recognized by the recognition means The suspended | floating matter detection apparatus of any one of Claim 1 to 3 provided with a means. Area specifying means for specifying an area having a luminance higher than a predetermined luminance from the water surface image;
An area calculating means for calculating the area of the area specified by the area specifying means,
2. The extraction unit according to claim 1, wherein the extraction unit extracts a region in which the area calculated by the area calculation unit matches a predetermined area condition from the regions specified by the region specification unit. 6. The suspended | floating matter detection apparatus of any one of 5. A perimeter calculation unit that calculates the perimeter of the region extracted as a region that matches the area condition,
The extraction unit extracts a region in which the perimeter calculated by the perimeter calculation unit matches a predetermined perimeter condition from regions extracted as regions that match the area condition. The suspended | floating matter detection apparatus of Claim 6. A laser light source for irradiating laser light;
The imaging means captures the reflected light of the laser light source as an image,
The floating object detection apparatus according to any one of claims 1 to 7, wherein a wavelength of laser light emitted by the laser light source is 1400 nm or more and 2600 nm or less. A suspended matter detection method for detecting suspended matter floating on water,
An imaging process for imaging the water surface;
An extraction step for extracting a region that matches the extraction condition of the suspended matter from the image of the water surface imaged by the imaging step;
A similarity calculation step of calculating a similarity between the upper region of the region extracted by the extraction step that is not easily affected by the state of waves around the suspended matter and a template image for detecting the suspended matter,
A detection step of detecting a region extracted by the extraction step as a region where the suspended matter exists when the similarity calculated by the similarity calculation step is higher than a predetermined value;
A suspended matter detection method comprising:

Images