JP2012026881A - Collapse detection system and collapse detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a collapse detection system accurately detecting a collapse that is a detection object by distinguishing it from a disturbance phenomenon.SOLUTION: A collapse detection system comprises: a photographing part 11 for photographing a monitored image of a detection object; an arithmetic operation part 13 for applying a time-space Gabor filter to an image photographed by the photographing part 11; a mask generating part 17 for generating a mask for excluding, as an exclusion area, a change area having equal to or more than a certain size among the image photographed by the photographing part 11; a mask processing part 14 for applying a mask processing to a processing result of the time-space Gabor filter using the mask generated by the mask generating part 17; and a determination part 15 for determining presence of a collapse using a processing result of the mask processing part 14.

Description

  The present invention relates to a collapse detection system and a collapse detection method for detecting a collapse such as a cliff collapse, and more particularly, to a collapse detection system and a collapse detection for detecting a small collapse phenomenon such as a pebbles falling as a precursor of a large-scale collapse. Regarding the method.

  In general, in monitoring of a landslide, a large rock collapse is detected using a wire sensor. In the work at the risk point of earth and sand disaster such as restoration work at the disaster site, it is possible to reduce the risk by discovering small collapses (hereinafter referred to as “small collapses”) such as pebbles falling as a precursor to a large collapse. It is also important to avoid it. Usually, a supervisor is assigned to visually monitor a dangerous spot. However, it is difficult for a human to cover a wide range, and monitoring by a machine is desired as a means for supporting a human.

  There are not many precedent examples of video surveillance specializing in earth and sand disasters, but considering moving pebbles as moving objects, various moving object detection technologies are being studied in the field of monitoring humans and vehicles. However, in the surveillance video, the small collapse is very small compared to the monitored area, and the color is similar to the background slope, so the normal moving object detection method identifies the small collapse as noise etc. There was a problem that it was difficult. As a technique to solve this problem, focusing on the fact that noise becomes an isolated point in the space-time where the frames of the monitoring video are arranged in time series, the collapse forms a continuous trajectory, There is a technique using a spatio-temporal Gabor filter characterized by improving the discrimination performance from noise (Non-patent Document 1).

  The spatiotemporal Gabor filter obtains local spatiotemporal frequency information of an input image by a convolution operation using a three-dimensional Gabor function in which a sine wave is localized using a Gaussian function as a window function. Usually, a plurality of types of spatiotemporal Gabor filters are prepared and a filter set is configured. By calculating the motion energy with all the spatiotemporal Gabor filters prepared, it is possible to obtain a response to the motion that causes various trajectories. Specifically, given that the spatio-temporal Gabor filter has a strong response when the corresponding parameter matches the moving direction / velocity of the object, Gabor for the speed, moving direction, etc. that is assumed to be small collapse. When the filter response (kinetic energy) exceeds a predetermined threshold, it is determined that a small collapse has occurred.

Ken Okamoto et al. "Detection of landslide disaster precursors using spatio-temporal Gabor filter" IEICE technical report. PRMU, Pattern recognition and media understanding Vol. 108, No. 327, pp.139-144, 2008.

  By the way, in video surveillance, in addition to identifying a detection target and noise included in the video, it is also necessary to identify a change (disturbance phenomenon) other than a collapse phenomenon such as flying insects. Small collapse is a phenomenon that occurs from the top to the bottom of the slope, and its direction can be assumed in advance, so a spatio-temporal Gabor filter that responds strongly to the vertical movement of the slope is used. If the response characteristic with respect to the moving direction is used, it is possible to distinguish a certain level of small collapse and disturbance phenomenon.

  However, the spatio-temporal Gabor filter response is affected not only by the movement direction (spatio-temporal characteristics) of the detection target but also by spatial characteristics such as the shape of the detection target and the contrast (brightness difference) with the background. . The small collapse can be regarded as a forward motion with a low contrast, but for example, when a bird with a high contrast with the background performs an orthogonal motion in the direction of the spatiotemporal Gabor filter, it is theoretically difficult to distinguish between the two. That is, the method based on the response characteristics of the spatio-temporal Gabor filter in the moving direction and speed cannot always distinguish between a disturbance phenomenon and a small collapse.

  An object of the present invention is to provide a collapse detection system and a collapse detection method that can accurately detect a collapse that is a detection target as a disturbance phenomenon, in view of the conventional technology described above.

  In order to achieve the above object, the invention according to the first aspect is a collapsing detection system for detecting collapsing, in which an imaging unit that captures a monitoring video to be detected, and an image captured by the imaging unit An arithmetic unit that applies a spatio-temporal Gabor filter, a mask generation unit that generates a mask for excluding a change area of a certain size or more from an image captured by the imaging unit, and the mask generation unit The gist is provided with a mask processing unit that performs mask processing on the processing result of the spatio-temporal Gabor filter using a mask, and a determination unit that determines the presence or absence of collapse using the processing result of the mask processing unit.

  The invention according to a second aspect is the invention according to the first aspect, wherein the mask generation unit extracts an inter-frame difference process in which a pixel whose luminance difference from the previous frame is larger than a predetermined threshold is extracted, and between the frames The gist of the invention is to generate the mask by performing a morphological process for extracting a change area of a certain size or larger and a peripheral area thereof on the processing result of the difference process.

  In order to achieve the above object, the invention according to the third aspect is a collapsing detection method for detecting collapsing, in which a photographing step for photographing a monitoring image to be detected and a video photographed in the photographing step are provided. A calculation step for applying a spatiotemporal Gabor filter to the mask, a mask generation step for generating a mask for excluding a change area of a certain size or more from the video imaged in the imaging step, and generation in the mask generation step The gist of the present invention includes a mask step for performing mask processing on the processing result of the spatio-temporal Gabor filter using the mask, and a determination step for determining the presence or absence of collapse using the processing result of the mask step.

  The invention according to a fourth aspect is the invention according to the third aspect, wherein in the mask generation step, an inter-frame difference process for extracting a pixel whose luminance difference from the previous frame is larger than a predetermined threshold, and the inter-frame difference The gist of the invention is to generate the mask by performing a morphological process for extracting a change area of a certain size or larger and a peripheral area thereof on the processing result of the difference process.

  According to the present invention, it is possible to provide a collapse detection system and a collapse detection method that can accurately detect a collapse that is a detection target by identifying it as a disturbance phenomenon.

It is a lineblock diagram of a collapse detection system in an embodiment of the invention. It is a flowchart which shows operation | movement of the collapse detection system in embodiment of this invention. It is explanatory drawing of the three-dimensional space in embodiment of this invention. It is a figure which shows the specific example of a process of the collapse detection system in embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  First, the outline of the present invention will be described. That is, in the present invention, inter-frame difference processing and morphological processing are performed in parallel with the spatio-temporal Gabor filter processing, a change area on the screen of a certain size or more is captured, and the change area and its surrounding areas are Disable filter response (kinetic energy). Originally, spatio-temporal Gabor filter processing is introduced to distinguish small collapses that are similar in color to the background and small in size to the same level as noise. For large changes in size, conventional video recognition processing is used. Can be easily captured and tracked. Therefore, in the present invention, the change that can be handled by the conventional method is excluded from the determination target by the spatiotemporal Gabor filter. In this way, it becomes possible to discriminate between disturbance phenomena such as the flight of a large bird whose speed selectivity of the spatio-temporal Gabor filter does not always function effectively and small collapse of the detection target.

  FIG. 1 is a configuration diagram of a collapse detection system 10 according to an embodiment of the present invention. This collapse detection system 10 is a system for detecting collapse such as a cliff collapse by video, and functionally, an imaging unit 11, a three-dimensional data creation unit 12, a calculation unit 13, a mask processing unit 14, and the like. The determination unit 15, the output unit 16, and the mask generation unit 17 are provided. The imaging unit 11 captures a monitoring image to be detected. The 3D data creation unit 12 creates 3D data using the video imaged by the imaging unit 11. The calculation unit 13 applies a spatio-temporal Gabor filter to the three-dimensional data created by the three-dimensional data creation unit 12 (video captured by the imaging unit 11). The mask generation unit 17 generates a mask for excluding a change area of a certain size or more from the video imaged by the imaging unit 11. Specifically, an inter-frame difference process that extracts pixels whose luminance difference from the previous frame is greater than a predetermined threshold, and a change area of a certain size or larger and its peripheral area with respect to the result of the inter-frame difference process. The extracted morphology process is performed to generate a mask. The mask processing unit 14 performs mask processing on the processing result of the spatio-temporal Gabor filter using the mask generated by the mask generation unit 17. The determination unit 15 determines the presence or absence of collapse using the processing result of the mask processing unit 14. The output unit 16 outputs the determination result of the determination unit 15 to a liquid crystal display or the like.

  FIG. 2 is a flowchart showing the operation of the collapse detection system 10 according to the embodiment of the present invention. Hereinafter, the configuration of the collapse detection system 10 will be described with reference to FIG.

  First, the monitoring video is converted into two-dimensional image × time three-dimensional data, and a spatio-temporal Gabor filter process is performed as a spatio-temporal feature extraction process of the three-dimensional data (steps S 1 → S 2 → S 3). . In parallel with this, mask generation processing is performed. Specifically, inter-frame difference processing and threshold processing are performed on consecutive frames of the monitoring video, and pixels with a certain change or more are extracted (steps S7 → S8). Thereafter, a morphological process is performed to extract a change area having a certain size or more and its peripheral area (step S9). After the area extracted by this process is set as an exclusion area, the spatiotemporal Gabor filter processing result is masked (step S4), and a small collapse determination process is performed (step S5).

The spatio-temporal Gabor filter process is a feature extraction process that captures the motion characteristics of the input image. Specifically, the input image (I) and a linear filter (g _φ ) are subjected to a convolution operation (represented by *) to react The movement energy (E) is obtained from the strength (r _φ ).

The linear filter (g _φ ) is expressed by the following formula.

Where γ is the size of the spatial scale, σ is the standard deviation of the Gaussian function applied to the spatial plane, v is the velocity representing the center position of the Gaussian function applied to the spatial plane, λ is the spatial period (1 / λ is the spatial frequency) ), Φ is the phase correction, μ _t is the average value of the Gaussian function applied in the time direction, τ is the standard deviation of the Gaussian function applied in the time direction, and θ represents the spatial direction of the filter. θ corresponds to the direction of collapse, and v corresponds to the speed of collapse. The linear filter is composed of parameters of direction (θ), velocity (v), and period (λ). For example, for the purpose of detecting downward collapse of about 4 × 4 pixels, settings such as θ = 0 ° (downward), v = 3 pixel / frame, and λ = 8 pixel are appropriate. Of course, other values can be used as parameters, and the detection characteristics can be changed by changing the parameters.

  FIG. 3 is an explanatory diagram of the three-dimensional space in the embodiment of the present invention. Here, an input image is represented in a three-dimensional space by a spatial pixel position (X-axis and Y-axis) and a temporal position (T-axis) of the pixel, and only pixels whose luminance values have changed on the image are displayed. it's shown. Changes in the image include not only collapse but also video compression noise, flicker noise, and the like. However, collapse causes a movement locus to appear linearly, and noise appears as an isolated point in time and space. Since the spatio-temporal Gabor filter exhibits high motion energy (E) with respect to a linear trajectory, the trajectory of collapse can be accurately detected from noise. Since the phenomenon and noise to be detected can be separated during the feature extraction process, it is not necessary to perform advanced pattern recognition processing in the judgment process, and it is possible to detect collapse with high accuracy even with simple threshold processing. Become.

  However, when the spatial characteristics change strongly, the spatio-temporal Gabor filter may not work effectively as shown in FIG. That is, (a) is a monitoring image when the bird 22 crosses the slope where the small collapse 21 occurs at a high speed. (B) is the result of applying a spatiotemporal Gabor filter that responds to vertical movement, and the white portions 21 and 22 are response points. Because the birds are large in size and have a large contrast with the background, the spatiotemporal Gabor filter responds to birds that move in a direction perpendicular to the direction to be extracted, resulting in greater kinetic energy than the small collapse of the detection target. It has been. It is possible to exclude responses to birds if the judgment threshold is increased, but small collapse cannot be detected. Therefore, it is difficult to discriminate between this disturbance phenomenon and small collapse by adjusting the normal judgment threshold. is there.

  Therefore, in the present invention, mask generation processing is performed in parallel with the spatiotemporal Gabor filter processing. That is, (c) is the inter-frame difference processing result binarized by threshold processing that is slightly higher than noise, and the black portion is the change region (pixel). (D) is obtained by subjecting the result to a morphological treatment. The morphological process is composed of a compression process for removing a change area of a certain size or less, and an expansion process for extracting a change area of a certain size or more and its surrounding area. (E) is the final result of applying the image of (d) to the spatiotemporal Gabor filter processing result (b) as a mask image excluding the changed portion 22.

  Usually, since the small collapse is a color similar to the background, the luminance difference from the background is smaller than the binarization threshold, and the small collapse area often does not remain in the inter-frame difference processing result as in the example (c). . As a result, in the mask image (d) after the morphological process, only the change accompanying the movement of the bird and its surrounding area are extracted. Accordingly, it is possible to effectively remove only the bird region by mask processing using the extraction region as an excluded region, that is, to extract a small collapse. On the other hand, even when the small collapse area 21 remains as in (c ′), the size of the change area associated with the small collapse is very small. Therefore, the small collapse region 21 is removed during the compression process of the morphological process and does not remain in the image (d) that becomes the mask image. Therefore, as in the case of (c), it is possible to remove only the bird region.

  As described above, according to the collapse detection system 10 in the embodiment of the present invention, it is possible to accurately detect a collapse that is a detection target by identifying it as a disturbance phenomenon. In other words, landslides and landslides can be automatically detected from the precursory stage by identifying them as disturbances such as flying birds, ensuring work safety at disaster recovery activities and other places where landslide disasters occur (disaster Prevention). Although false alarms are a problem in these automatic detection systems, the present invention solves the problem of discrimination from disturbance phenomenon and contributes to improvement of practicality. As a function of the disaster prevention system, early detection in sediment disaster monitoring becomes possible by combining it with the function of automatically sending alarms such as alarms and e-mail delivery to the monitoring center and disaster prevention manager from the judgment result of collapse occurrence.

  In FIG. 4, the identification of bird flight has been described as an example, but this also works effectively for phenomena such as tree shaking. In addition, the present invention is also effective for general movement detection of microscopic objects having similar colors to the background environment similar to the landslide disaster precursor. For example, it can be applied to monitoring of living things in a low-contrast environment where illumination is not available.

DESCRIPTION OF SYMBOLS 10 ... Fall detection system 11 ... Image pick-up part 12 ... Three-dimensional data creation part 13 ... Calculation part 14 ... Mask processing part 15 ... Determination part 16 ... Output part 17 ... Mask generation part

Claims (4)

A collapse detection system for detecting a collapse,
A shooting section for shooting the surveillance video to be detected;
An arithmetic unit that applies a spatiotemporal Gabor filter to the video imaged by the imaging unit;
A mask generation unit that generates a mask for excluding a change area of a certain size or more from the video imaged by the imaging unit;
A mask processing unit that performs mask processing on the processing result of the spatio-temporal Gabor filter using the mask generated by the mask generation unit;
A determination unit that determines the presence or absence of collapse using the processing result of the mask processing unit;
A collapse detection system characterized by comprising:   The mask generation unit includes an inter-frame difference process that extracts pixels having a luminance difference from a previous frame that is greater than a predetermined threshold, a change area that is a predetermined size or larger with respect to a process result of the inter-frame difference process, and its peripheral area The collapse detection system according to claim 1, wherein the mask is generated by performing a morphological process of extracting a mask. A collapse detection method for detecting collapse,
A shooting step for shooting the surveillance video to be detected;
A calculation step of applying a spatiotemporal Gabor filter to the video shot in the shooting step;
A mask generation step of generating a mask for excluding a change area of a certain size or more from the video imaged in the imaging step;
A mask step of performing mask processing on the processing result of the spatiotemporal Gabor filter using the mask generated in the mask generation step;
A determination step of determining the presence or absence of collapse using the processing result of the mask step;
A collapse detection method characterized by comprising:   In the mask generation step, an inter-frame difference process for extracting a pixel whose luminance difference from the previous frame is larger than a predetermined threshold, and a change area having a predetermined size or more with respect to a process result of the inter-frame difference process and its peripheral area The collapse detection method according to claim 3, wherein the mask is generated by performing a morphological process of extracting the mask.