Snow accumulation detection system

Abstract

To provide a snow accumulation detection system that is adaptable to various types of installation environments and enables snow accumulation determination with improved detection accuracy.SOLUTION: A snow accumulation detection system comprises: a target installed in a monitoring range; and a controller for performing snow accumulation detection on the basis of an image obtained by imaging the target using a camera. The controller: acquires a reference image at a time of non-snow accumulation; sequentially acquires detected images according to a first detection interval during an inspection period; calculates a difference image between a binarized reference image after binarization processing and a binarized detected image after binarization processing; with respect to respective pixels of the difference image, calculates, as a snow accumulation score, the number of pixels at which a difference exceeding an allowable difference amount is calculated; in response to the sequentially acquired detected images, sequentially calculates snow accumulation scores; and determines that a present state is a snow accumulation state and outputs a snow accumulation detection signal in an ON state in the case that in the present state the sequentially calculated snow accumulation scores equal to or larger than a previously set first threshold continue the previously set number of continuous times.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a snow accumulation detection system for detecting the presence or absence of snow accumulation in a monitoring area by utilizing image processing technology. [Background technology]

[0002] Regarding snow accumulation determination, there are various methods of detecting snow accumulation as conventional techniques to avoid the problems of existing humidity sensors or temperature sensors. In particular, in recent years, a device has been proposed that captures the road surface condition with a camera installed outdoors, performs some kind of image processing on the captured image, and then determines whether there is snow accumulation or not (for example, see Patent Documents 1 and 2). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2021-161736 A [Patent Document 2] JP 2005-308437 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, the conventional techniques disclosed in Patent Documents 1 and 2 have the following problems. Patent Document 1 relates to a road heating switch control system. In this system, an image captured by a camera installed outdoors is compared with an image captured in advance when there is no snow, after noise removal and binarization processing, to determine the state of snow.

[0005] In this way, the system according to Patent Document 1 judges snow accumulation by comparing a reference image (reference image) with an image to be detected (detection image). However, this system is designed to control road heating installed in an environment with a limited number of users, such as a parking lot for an ordinary home, and does not assume an environment where an unspecified number of vehicles and people pass through the target area.

[0006] Furthermore, the system disclosed in Patent Document 1 judges differences across the entire imaging range, which may increase the risk of false positives if an obstacle such as a vehicle or person remains within the imaging range for a certain period of time.

[0007] In addition, the system disclosed in Patent Document 1 sets the standard for ending snow accumulation detection as the time from the start of detection, which may result in the snow accumulation detection output ending before the snow has completely melted.

[0008] Patent Document 2 relates to a snow detection system. In the system according to Patent Document 2, existing road markings such as center lines are imaged as targets and used as markings for judgment criteria in detecting snow accumulation.

[0009] In the system of Patent Document 2, for example, the lines and the like that serve as targets in a parking lot are limited to lines separating parking spaces and signs such as "Stop" and "Stop temporarily". These signs are frequently blocked by parked vehicles and users, and are also blocked by parking for long periods of time. Therefore, the system of Patent Document 2 has a problem in that it cannot capture an image of the target while it is blocked.

[0010] As described above, both of the systems disclosed in Patent Documents 1 and 2 have restrictions on the installation environment, and in various installation environments, there is a risk that the accuracy of snow accumulation detection will deteriorate or snow accumulation detection itself will not be possible.

[0011] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a snow accumulation detection system that can be adapted to various installation environments and enables snow accumulation determination with improved detection accuracy. [Means for solving the problem]

[0012] The snow accumulation detection system according to the present disclosure includes a target provided in a monitoring range in which the snow accumulation state is to be detected, a camera that captures an image of the target and outputs an image of the target, and a control device that detects snow accumulation based on an image acquired from the camera, the target being positionable by edge extraction processing through image processing and having a pattern that allows an amount of snow accumulation on the target to be estimated by binarization processing through image processing, the control device sets an image acquired from the camera when there is no snow accumulation as a reference image, and sets images acquired from the camera in time series in accordance with a preset first detection interval as detection images during an inspection to determine the snow accumulation state, and detects snow accumulation based on the reference image in the area of ​​the target. With regard to the detected image in the target area, an image after binarization processing based on a preset brightness range is generated as a binary reference image, and with regard to the detected image in the target area, an image after binarization processing based on the brightness range is generated as a binary detection image, a difference image between the binary reference image and the binary detection image is calculated, and the number of pixels for which a difference exceeding a preset allowable difference amount is calculated as a snow accumulation score for each pixel of the difference image is calculated as a snow accumulation score, and snow accumulation scores are calculated sequentially corresponding to the detected images acquired sequentially in time series, and if a state in which the sequentially calculated snow accumulation scores are equal to or greater than a preset first threshold value continues for a preset number of consecutive times, it is determined that a snow accumulation state exists, and a snow accumulation detection signal is output as an ON state. Effect of the Invention

[0013] According to the present disclosure, it is possible to obtain a snow accumulation detection system that can be adapted to various installation environments and enables snow accumulation determination with improved detection accuracy. [Brief description of the drawings]

[0014] [Figure 1] 1 is an explanatory diagram illustrating an overall configuration of a snow accumulation detection system according to a first embodiment of the present disclosure. [Diagram 2] FIG. 2 is an explanatory diagram showing the positional relationship of each component when the snow accumulation detection system is applied to a pay-by-the-hour parking lot in the first embodiment of the present disclosure. [Diagram 3] FIG. 2 is an explanatory diagram showing a specific example of the shape of the target according to the first embodiment of the present disclosure. [Figure 4] 4 is a flowchart illustrating a snow accumulation detection process executed in the snow accumulation detection system according to the first embodiment of the present disclosure. [Diagram 5] FIG. 4 is an explanatory diagram relating to pre-processing executed by the control device in the first embodiment of the present disclosure. [Figure 6] 2 is an explanatory diagram showing the luminance distribution of each element included in a multi-valued image obtained by a camera in the first embodiment of the present disclosure. FIG. [Figure 7] FIG. 11 is an explanatory diagram showing a specific example when the control device determines YES in the process of step S408 in the first embodiment of the present disclosure. [Figure 8] FIG. 11 is an explanatory diagram showing a specific example when the control device determines NO in the process of step S408 in the first embodiment of the present disclosure. [Figure 9] FIG. 2 is an explanatory diagram relating to a case where a snow accumulation score is equal to or greater than a threshold and a case where the snow accumulation score is less than the threshold in the first embodiment of the present disclosure. [Figure 10] 4 is a flowchart illustrating a process of detecting no snow accumulation executed in the snow accumulation detection system according to the first embodiment of the present disclosure. [Figure 11] 11 is a flowchart of a portion related to a no-snow detection process executed in the snow detection system according to the first embodiment of the present disclosure, and further related to a delay timer process executed. [Figure 12] FIG. 2 is a diagram illustrating an operation model of the snow accumulation detection system according to the first embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] Hereinafter, preferred embodiments of the snow accumulation detection system of the present disclosure will be described with reference to the drawings. The snow detection system disclosed herein uses a target suitable for determining whether snow is present or not through image processing, and quantitatively determines whether snow is present or not based on the results of a comparison between a reference image of the target acquired when there is no snow and a detection image of the target acquired when detecting snow-covered and non-snow-covered conditions, and has a technical feature in that various parameters used when making the comparative determination through image processing can be appropriately set according to the installation environment.

[0016] Before providing a detailed description of the snow accumulation detection system according to the present disclosure, the definitions of technical terms used in the first embodiment will be described below.

[0017] Reference image: An image taken in advance of the target to be used for comparison when there is no snow cover. Detection image: An image of a target captured during snowfall detection, used to compare with a reference image when determining whether snowfall has been detected. Note that the detection image can be a still image captured at each sampling period, and can also be captured as a still image at set intervals from a video captured in real time.

[0018] Detection Interval: A set time that dictates when to capture a detection image. Snow cover score: The difference between the detected image and the reference image after preprocessing such as edge extraction and binarization.

[0019] Snow accumulation determination: When the snow accumulation score is equal to or greater than a threshold, it is determined that snow was present at the time the detection image was captured. Number of judgments: The number of consecutive times that snow accumulation has been judged to be snow accumulation. Snow accumulation detection: When the set number of snow accumulation judgments have been made, it is determined that snow has accumulated and the snow accumulation detection signal is turned ON.

[0020] Embodiment 1 1 is an explanatory diagram showing an overall configuration of a snow accumulation detection system according to a first embodiment of the present disclosure. A snow accumulation detection system 10 in the first embodiment includes a target 1, a camera 11, and a control device 12.

[0021] The camera 11 captures an image including the target 1. The control device 12 performs a series of processes for detecting snow accumulation on the area of ​​the target 1 included in the image captured by the camera 11, and when detecting snow accumulation, outputs a snow accumulation detection signal to an external device such as the snow melting device 20 in an ON state.

[0022] The target 1 is a determination pattern drawn or placed in a monitoring area where the snow accumulation condition should be detected in order to determine the snow accumulation condition. The control device 12 detects the snow accumulation condition of the target 1 by performing image processing on the image including the target 1 captured by the camera 11, focusing on the target 1 portion.

[0023] The target 1 will be described in detail later with reference to Fig. 2 and Fig. 3. A specific series of processes for detecting snow accumulation will be described later with reference to Figs.

[0024] In the following, a specific example will be described in which the snow accumulation detection system according to the present disclosure is applied to a pay-by-the-hour parking lot. In the following, a specific example will be described in which a real-time video image is captured by the camera 11, a still image is acquired at a preset detection interval by the control device 12, and image processing is performed on each still image.

[0025] <Layout Description> A specific layout when the snow accumulation detection system according to the present disclosure is applied to a pay-by-the-hour parking lot will be described with reference to Figures 2 and 3. Figure 2 is an explanatory diagram showing the positional relationship of each component when the snow accumulation detection system is applied to a pay-by-the-hour parking lot in the first embodiment of the present disclosure.

[0026] Figure 2(A) shows an example of the placement of target 1 in a pay-by-hour parking lot, and Figure 2(B) shows the positional relationship between target 1 drawn in a state corresponding to the monitoring area of ​​the pay-by-hour parking lot and a camera 11 installed in the pay-by-hour parking lot.

[0027] 3 is an explanatory diagram showing a specific example of the shape of the target 1 according to the first embodiment of the present disclosure. The snow accumulation detection system 10 detects snow accumulation based on the state of snow accumulated on the target 1 installed in a pay-by-the-hour parking lot. The target 1, the camera 11, and the control device 12 will be described in detail with reference to FIGS. 2 and 3.

[0028] [1] Target 1 Description The target 1 is installed in a monitoring range where the snow accumulation condition is to be detected, for example, on the road surface of a pay-by-the-hour parking lot, in order to determine the snow accumulation condition based on image processing. The number of targets 1 installed may be more than one depending on the installation environment. Figure 2(A) shows an example in which targets 1 are installed in two locations in order to detect snow accumulation in a pay-by-the-hour parking lot with 18 parking spaces.

[0029] Target 1 needs to be a certain size to reduce weather-related disturbances such as camera shaking due to wind. Also, it is important for target 1 to have a color and shape that provides good contrast with the road surface when binarized and makes it easy to identify the position in order to achieve high detection accuracy.

[0030] In other words, the target 1 is formed as a pattern that allows the position to be specified by edge extraction processing in image processing, and allows the amount of snowfall on the target 1 to be estimated by binarization processing in image processing. The edge extraction processing and binarization processing will be described later.

[0031] In the present embodiment 1, the specific shape of the target 1 is as shown in Fig. 3, with an outer shape of width L1 (mm), length L2 (mm), and line width of width ΔL1 (mm), length ΔL2 (mm). When drawing this target 1 in a part of the hourly parking lot, the diagonally hatched part within the outer shape in Fig. 3 (part marked "A part" in Fig. 3) corresponds to the part where the road surface of the hourly parking lot is exposed, and the other part (part marked "B part" in Fig. 3) corresponds to the marking part to be drawn.

[0032] A specific example is a target 1 in which L1 = L2 = 2000 mm, ΔL1 = ΔL2 = 200 mm, and the marking portion is white.

[0033] Target 1 may be installed anywhere along the path of vehicles, but it is preferable to avoid installing it at locations where vehicles frequently pass by, such as intersections, or near parking lots, fuel trucks, garbage trucks, security company vehicles, or other work vehicles stopping or working spaces.

[0034] Furthermore, as the target 1, a sticker-type road marker that can be easily installed can also be used as long as durability, weather resistance, and the like are ensured and the size is such that it can be captured as a camera image.

[0035] Note that if there are no conditions that prevent imaging, such as frequent obstruction by parked vehicles or users, or long periods of parking, and the contrast between parts A and B can be obtained as shown in Figure 3, it is also possible to use road markings already on the road surface as target 1.

[0036] [2] Explanation of the installation position of camera 11 2(B), the camera 11 can be installed on a pole 2 on which a housing for accommodating the control device 12 is installed. The camera 11 is installed in a position that does not interfere with the parking space, vehicle paths, or pedestrian (user) paths, and can capture an image of the target 1.

[0037] If a building or an overhead pole can be used instead, the camera 11 may be installed by utilizing the existing facility without installing a new pole. Also, multiple cameras 11 may be installed according to the target of snow accumulation detection.

[0038] [3] Camera 11 Description The camera 11 in the present embodiment 1 is assumed to be a network camera capable of transmitting video to the control device 12. The camera 11 is assumed to be installed outdoors during the winter, and is waterproof and dustproof so that it can function without any problems even in a low-temperature environment.

[0039] Furthermore, the camera 11 is capable of capturing images day and night, and is capable of stably capturing images of the target 1 even in bad weather such as snow.

[0040] [4] Description of control device 12 The control device 12 is made up of a PC, which is the main body of the control device, a PoE HUB that connects the PC and the network camera, and a constant-voltage power supply unit that supplies power, and is housed in a weather-resistant housing. As with the camera 11, the control device 12 is one that can operate without problems even in low-temperature winter environments. In other words, it is desirable for the control device 12 to operate with the same degree of accuracy as the camera 11.

[0041] The control device 12 according to the first embodiment has the following functions. A function that captures still images at regular intervals from the captured video to obtain detection images. A function for performing edge extraction and binarization as preprocessing on images captured by the camera 11. After pre-processing, the function compares the difference between the binary images of the reference image and the detected image, and counts the number of pixels with a difference that exceeds a pre-set allowable difference amount as the snow accumulation score. - A function that outputs a snow accumulation detection signal based on the results of comparing the snow accumulation score with a threshold value. - A function to determine whether snow accumulation detection has been cancelled.

[0042] Furthermore, the control device 12 according to the first embodiment also has the following additional functions. Additional function 1: A function that can perform judgment processing using multiple cameras and multiple targets can be further provided. When performing judgment processing using multiple targets, it is optional to decide whether snow accumulation detection is to be determined based on snow accumulation judgment for all targets 1, or based on snow accumulation judgment for any one of targets 1. In other words, it is possible to select whether the snow accumulation detection for each target 1 is based on an AND condition or an OR condition.

[0043] Additional function 2: When snow accumulation is detected, a snow accumulation detection signal is output to the snow melting device 20, and at the same time, a function to notify a pre-registered manager, etc. by email or the like that the snow melting device 20 will be activated, and to encourage snow removal, can be added. Also, a function to notify necessary equipment status such as equipment abnormalities can be added.

[0044] Next, a series of processes executed in the snow accumulation detection system according to the first embodiment will be described in detail with reference to FIGS.

[0045] 4 is a flowchart related to a snow accumulation detection process executed in the snow accumulation detection system according to the first embodiment of the present disclosure. The control device 12 executes a series of processes from step S401 to step S409 shown in FIG. 4, thereby quantitatively determining whether or not to output an ON-state of the snow accumulation detection signal.

[0046] In step S401, the control device 12 acquires an image during normal times, which corresponds to a time when there is no snowfall, as a reference image from the video captured by the camera 11 in real time.

[0047] Next, in step S402, the control device 12 initializes to 0 a determination count n for counting the number of times snow has been detected in a snow accumulation determination, which will be described later.

[0048] Next, in step S403, the control device 12 acquires, as a detection image, an image to be used for comparison with the reference image from the video captured by the camera 11 in real time during the inspection when snow accumulation determination is to be performed.

[0049] The acquisition of the detection images is repeated periodically at a preset first detection interval when the process returns to step S403 from step S408 described later. That is, during an inspection for performing the snow accumulation detection process, the detection images are acquired in chronological order from the camera 11 at the preset first detection interval.

[0050] Next, in step S404, the control device 12 performs pre-processing on both the reference image acquired in step S401 and the detection image acquired in step S403 in preparation for the comparison process.

[0051] An example of the pre-processing executed by the control device 12 is to execute noise removal processing → edge extraction processing → binarization processing in this order on the image. Note that the conventional techniques disclosed in Patent Document 1 and the like can be applied to these pre-processing processes, and detailed explanations will be omitted.

[0052] Fig. 5 is an explanatory diagram relating to the pre-processing executed by the control device 12 in the first embodiment of the present disclosure. Fig. 5(A) shows a multi-value image before pre-processing is executed for each of the reference image and the detection image, and Fig. 5(B) shows a binary image after pre-processing is executed for each of the reference image and the detection image.

[0053] An image obtained by performing the binarization process on the reference image is generated as a binarized reference image, and an image obtained by performing the binarization process on the detected image is generated as a binarized detected image.

[0054] By executing a series of pre-processing steps on the target 1 shown in FIG. 3 above, the control device 12 can generate a binary image in which the outline of the marking portion, part B, is white and the other parts are black, as shown in FIG. 5(B).

[0055] A supplementary explanation will now be given regarding the binarization process, which is one of the pre-processing processes. Fig. 6 is an explanatory diagram showing the luminance distribution of each element included in the multi-valued image obtained by camera 11 in the first embodiment of the present disclosure. The area captured by camera 11 includes elements such as cars, headlights, shadows caused by sunlight, reflected light, and asphalt in addition to snow (blizzard, snow accumulation, and accumulated snow) that is originally desired to be detected.

[0056] In Figure 6, the horizontal axis represents 256 levels of brightness, and the brightness distribution of each element is shown. To improve the accuracy of snowfall detection, it is important to distinguish and recognize the following elements from snow: cars, headlights, shadows from sunlight, reflected light, and asphalt, as these can all cause false positives or over-detections.

[0057] 6, snow has a luminance distribution in the range of approximately 100 to 175, but headlights, shadows caused by sunlight, and reflected light have luminance distributions in the ranges of less than 100 or greater than 190, showing a luminance distribution different from that of snow. Therefore, the control device 12 can extract snow while excluding headlights, shadows caused by sunlight, and reflected light by performing binarization processing on the luminance range of 100 to 175 and other ranges.

[0058] This brightness distribution will differ depending on the area being monitored for snowfall (i.e., the environment in which the snowfall detection system is installed), but by measuring the brightness distribution for each element according to the area being monitored and setting a brightness range for performing appropriate binarization processing based on the measurement results, the accuracy of snowfall detection can be improved.

[0059] In the brightness distribution shown in Fig. 6, the brightness of cars, which are expected to frequently enter the imaging area, overlaps with the brightness of snow. It is difficult to suppress the influence of such cars and improve the accuracy of snowfall detection by binarization processing alone, but a method for suppressing the influence of cars will be described later using step S408 and Figs. 7 and 8.

[0060] Next, in step S405, the control device 12 performs a comparison process on both the binarized reference image and the binarized detection image binarized after the preprocessing in step S404. Specifically, the control device 12 calculates the difference in luminance between the binarized reference image and the binarized detection image for each pixel, and generates a difference image.

[0061] Furthermore, the control device 12 calculates, for each pixel of the difference image, the number of pixels for which a difference exceeding a preset allowable difference amount is calculated as a snow accumulation score.

[0062] Next, in step S406, the control device 12 judges whether the snow accumulation score calculated based on the comparison result in step S405 is equal to or greater than a first threshold value set in advance. If the judgment result in step S406 is YES, the control device 12 proceeds to processing in step S407. On the other hand, if the judgment result in step S406 is NO, the control device 12 returns to processing in step S402 and repeats the processing in and after step S402.

[0063] The first threshold value used in the comparison process with the snow accumulation score in the snow accumulation detection process can be freely changed according to the installation environment of the snow accumulation detection system. By setting an appropriate first threshold value according to the installation environment, it is possible to deal with various factors in the installation environment (such as differences in snow accumulation conditions depending on the region, differences in the use of the installation location such as a business office, hourly parking lot, or shared space in an apartment complex) or various output destination devices.

[0064] When the process proceeds to step S407, the control device 12 increments the number of determinations n to n + 1. As a result, the number of times that the snow accumulation score has consecutively been equal to or greater than the first threshold value is calculated as the number of determinations n.

[0065] Next, in step S408, the control device 12 judges whether the judgment number n incremented in step S407 has reached a preset continuous number N. If the judgment result in step S408 is YES, the control device 12 proceeds to processing in step S409. On the other hand, if the judgment result in step S408 is NO, the control device 12 returns to the processing in step S403 and repeats the processing in and after step S403.

[0066] In other words, the control device 12 sequentially calculates snow accumulation scores corresponding to detection images acquired sequentially in time series at the first detection interval, and repeats the process of determining whether the sequentially calculated snow accumulation scores are greater than or equal to the first threshold value.

[0067] Here, the process in step S408 will be supplementarily described with reference to Fig. 7 and Fig. 8. Fig. 7 is an explanatory diagram showing a specific example of a case where the process in step S408 by the control device 12 in the first embodiment of the present disclosure is judged as YES. Meanwhile, Fig. 8 is an explanatory diagram showing a specific example of a case where the process in step S408 by the control device 12 in the first embodiment of the present disclosure is judged as NO.

[0068] In both of the specific examples of FIG. 7 and FIG. 8, the parameters are set to the following values. First detection interval when acquiring a detection image in step S403: 10 seconds First threshold used for comparison with the snow cover score in step S406: 20 The number of consecutive times N used for comparison with the number of judgments n in step S408: 10

[0069] In Fig. 7, the lowest snow accumulation score calculated for the most recent 10 detection images is 28, and the state in which a snow accumulation score equal to or greater than the first threshold has been obtained 10 consecutive times is shown. This means that all 10 detection images acquired sequentially at 10-second intervals were determined to have snow accumulation, and snow accumulation was determined to have been obtained continuously for 90 seconds.

[0070] In such a case, the control device 12 determines in step S408 that accumulated snow has been detected, and proceeds to step S409.

[0071] On the other hand, Fig. 8 shows that the lowest snow accumulation score calculated for the most recent 10 detection images was 4, and that the snow accumulation score did not reach the first threshold value or more for 10 consecutive images. In other words, this means that none of the 10 detection images acquired sequentially at 10-second intervals was determined to have snow accumulation, and no snow accumulation determination was made for 90 consecutive seconds.

[0072] In such a case, the control device 12 determines in step S408 that snow accumulation is not detected. More specifically, when the snow accumulation score becomes less than the first threshold value in step S406, the control device 12 returns to the process of step S402, sets the number of determinations n to 0, and then repeats the series of processes.

[0073] For example, when a car is present on the target 1, the snow accumulation score becomes equal to or greater than the first threshold, as shown in the last detected image in Fig. 8. However, the control device 12 executes the process in step S408 so as not to proceed to step S409 unless the snow accumulation score becomes equal to or greater than the first threshold N consecutive times.

[0074] Therefore, by setting the number of consecutive occurrences N and the first detection interval to appropriate values ​​according to the installation environment, it is possible to prevent snow accumulation from being detected due to a car, person, or the like passing over the target 1.

[0075] 7 and 8, the process does not proceed to step S409 unless the snow accumulation scores calculated for the detection images acquired at 10-second intervals for 90 seconds are equal to or greater than the first threshold value. Therefore, unless a car, person, or the like remains on the target 1 for 90 seconds, snow accumulation detection due to a car, person, or the like can be prevented.

[0076] Here, the snow accumulation score will be further described with reference to Fig. 9. Fig. 9 is an explanatory diagram relating to a case where the snow accumulation score is equal to or greater than the first threshold value and a case where the snow accumulation score is less than the first threshold value in the first embodiment of the present disclosure.

[0077] 9(A) shows a state in which snow has accumulated on the target 1, the control device 12 has calculated the snow accumulation score to be 47, and the presence of snow has been detected. Also, in FIG. 9(B) shows a state in which the amount of snow on the target 1 has decreased as a result of a car passing over the target 1, but the control device 12 has calculated the snow accumulation score to be 27, and the presence of snow has been detected.

[0078] On the other hand, in Figure 9 (C), as a car passes over target 1, the amount of snow on target 1 decreases further than in the state of Figure 9 (B), and the control device 12 calculates the snow accumulation score to 17, which is below the first threshold, indicating that no snow has been detected.

[0079] In this way, the snow accumulation detection system according to the first embodiment can detect snow accumulation while taking into consideration that the amount of snow accumulation on the target 1 decreases as a result of cars, people, and the like passing by.

[0080] Returning to the explanation of Fig. 4, when the process proceeds to step S409, the control device 12 makes a final determination that snow accumulation has been detected after the snow accumulation determination has been made continuously for the set number of consecutive times N, and turns on the snow accumulation detection signal, which is output to an external device such as the snow melting device 20. As a result, by starting the snow melting device 20 after snow accumulation has been detected, the amount of snow accumulated in the hourly parking lot can be reduced.

[0081] In other words, the snow melting device 20 does not need to be kept running all the time, but can be started at the appropriate time based on the highly accurate snow accumulation detection results. By using the snow accumulation detection system of this embodiment 1, the start-up timing of external devices such as the snow melting device 20 can be optimized, thereby achieving power savings.

[0082] When the snow detection signal is finally output as ON in step S409 by the snow detection process described with reference to Fig. 4, the no-snow detection process is continued. Fig. 10 is a flowchart related to the no-snow detection process executed in the snow detection system according to the first embodiment of the present disclosure.

[0083] The control device 12 executes a series of processes from step S1001 to step S1006 shown in FIG. 10, thereby making it possible to quantitatively determine whether or not the snow accumulation detection signal that has been output as ON should be output as OFF.

[0084] In step S1001, the control device 12 acquires an image to be used for comparison with a reference image from a video captured in real time by the camera 11 as a detection image in order to detect the appropriate timing for switching the snow accumulation detection signal, which is in the ON state, to the OFF state.

[0085] The acquisition of the detection images is repeated periodically at a preset second detection interval when the process returns to step S1001 from step S1004 described later. That is, during an inspection for performing the no-snow detection process, the detection images are acquired from the camera 11 in chronological order at the preset second detection interval.

[0086] Next, in step S1002, the control device 12 performs pre-processing in preparation for a comparison process on both the reference image acquired in step S401 in Fig. 4 and the detection image acquired in step S1001. This pre-processing is the same as the process in the previous step S404, and a description thereof will be omitted.

[0087] Next, in step S1003, the control device 12 performs a comparison process on both the reference image and the detected image binarized after the preprocessing in step S1002, and finally calculates a snow cover score. This comparison process is the same as the process in the previous step S405, and therefore a description thereof will be omitted.

[0088] Next, in step S1004, the control device 12 judges whether the snow accumulation score calculated based on the comparison result in step S1003 is less than a preset second threshold value. If the judgment result in step S1004 is YES, the control device 12 proceeds to the process of step S1005. On the other hand, if the judgment result in step S1004 is NO, the control device 12 returns to the process of step S1001 and repeats the processes from step S1001 onward.

[0089] If the process proceeds to step S1005, the control device 12 determines whether to immediately turn the snow accumulation detection signal, which is in the ON state, to the OFF state, or to turn it to the OFF state after a preset output delay time M (seconds) has elapsed.

[0090] A supplementary explanation of the reason for setting the output delay time M is provided below. In the actual operation of the snow melting device 20, etc., there may be a situation where it is desired to output a contact beyond the time for snow accumulation detection and continue operation for a certain period of time. When snow continues to fall to a degree that does not trigger snow accumulation detection, it may be desired to operate the snow melting device 20 until the snow stops completely, or when the snow melting device 20 uses an electric heating wire method or the like and cools down as soon as it stops operating, causing residual snow to remain.

[0091] To accommodate these cases, the output delay time M can be variably set to a value of 0 seconds or more so that an extension time can be set for the snow accumulation detection signal output. When setting the output extension time M, the installation environment conditions should be carefully examined and the minimum extension time required should be set. In installation environments where there is no need to extend the snow accumulation detection, the output extension time M can be set to 0 seconds, thereby eliminating the need for the output extension time M.

[0092] Therefore, in step S1005, if the output delay time M is set to 0, the control device 12 determines not to perform the output delay process (i.e., NO), and proceeds to the process of step S1006. On the other hand, if the output delay time M is set as a positive real number, the control device 12 determines to perform the output delay process (i.e., YES), and proceeds to the process of step S1101 in FIG. 11, which will be described later.

[0093] When the process proceeds to step S1006, the control device 12 immediately switches the snow accumulation detection signal, which is in the ON state, to the OFF state, and ends the no-snow accumulation detection process.

[0094] 11 is a flowchart of a part related to the no-snow detection process executed in the snow detection system according to the first embodiment of the present disclosure, and further related to the output delay process executed. When the process proceeds to step S1101, the series of output delay processes shown in FIG. 11 are executed.

[0095] In step S1101, the control device 12 starts an output delay timer and commences output delay processing.

[0096] Next, in step S1102, the control device 12 acquires an image from the video captured by the camera 11 in real time as a detection image to be used for comparison with the reference image in order to monitor changes in the snow accumulation conditions while the output delay processing is being executed.

[0097] The acquisition of the detection images is repeated periodically at a preset second detection interval when the process returns to step S1102 from step S1106 described later. That is, in step S1102 of the output delay process shown in Fig. 11, the detection images are acquired in time series from the camera 11 at the preset second detection interval, similar to step S1001 of the no-snow detection process shown in Fig. 10.

[0098] Next, in step S1103, the control device 12 performs pre-processing in preparation for a comparison process on both the reference image acquired in step S401 in Fig. 4 and the detection image acquired in step S1102. This pre-processing is the same as the process in the previous step S404, and a description thereof will be omitted.

[0099] Next, in step S1104, the control device 12 performs a comparison process on both the reference image and the detected image binarized after the preprocessing in step S1103, and finally calculates the snow cover score. This comparison process is the same as the process in the previous step S405, and therefore a description thereof will be omitted.

[0100] Next, in step S1105, the control device 12 determines whether or not the snow accumulation score calculated based on the comparison result in step S1104 is equal to or greater than the first threshold value used in the snow accumulation detection process shown in Fig. 4. That is, in this step S1105, the control device 12 determines whether or not the detection state has changed from a state where there is no snow accumulation to a state where there is snow accumulation during the execution of the output delay process.

[0101] If the determination result in step S1105 is NO, i.e., if the control device 12 determines that the detection state of no snow is continuing, the control device 12 proceeds to the process of step S1106. On the other hand, if the determination result in step S1105 is YES, i.e., if the control device 12 determines that the detection state has changed from the detection state of no snow to the detection state of snow, the control device 12 proceeds to the process of step S1107.

[0102] In step S1106, the control device 12 judges whether the elapsed time of the output delay timer started in step S1101 has reached the output delay time M seconds. If the judgment result in step S1106 is YES, the control device 12 proceeds to the process of step S1006 in Fig. 10, turns the snow accumulation detection signal that is in the ON state to the OFF state after the output delay time M has elapsed, and ends the no-snow accumulation detection process.

[0103] On the other hand, if the determination result in step S1107 is NO, the control device 12 returns to the process of step S1102 and repeats the processes from step S1102 onwards.

[0104] Also, if the process proceeds to step S1107 based on the judgment result of step S1105, the control device 12 stops the output delay timer because the detection state has changed from no snow to the presence of snow while executing the output delay process, and sets the number of judgments n to 1. After that, the control device 12 proceeds to the process of step S403 in the snow detection process shown in Figure 4 above.

[0105] That is, in this step S1107, the control device 12 determines that the first state in which the snow accumulation score is equal to or greater than the first threshold has occurred, sets n to 1, stops the output delay timer, and then executes the processing from step S403 onwards, thereby transitioning from the no-snow accumulation detection processing to the snow accumulation detection processing.

[0106] As described above, by executing the series of processes shown in Figures 10 and 11, the control device 12 can determine whether to immediately turn the snow accumulation detection signal that is in the ON state OFF or to turn it OFF after the output delay time M (seconds) has elapsed, and can transition from the no snow accumulation detection process to the snow accumulation detection process when the detection state changes from a no snow accumulation detection state to a snow accumulation detection state.

[0107] In particular, the control device 12 can turn the snow accumulation detection signal from the ON state to the OFF state after the lapse of an output delay time M that is preset as a value equal to or greater than 0, and can stop the snow melting device 20 at an appropriate timing based on the highly accurate snow accumulation detection result. As a result, by using the snow accumulation detection system according to the first embodiment, the stop timing of external devices such as the snow melting device 20 can be optimized, and power saving can be achieved.

[0108] In the series of processes for no snow detection in Fig. 10, the snow condition in target 1 is quantitatively calculated as a snow accumulation score, and it can be accurately determined that the snow accumulation condition has been eliminated when the snow accumulation score becomes less than the second threshold. Therefore, in the process for no snow detection shown in Fig. 10, unlike the process for snow detection shown in Fig. 4, multiple determinations are not made, and it is determined that there is no snow accumulation when the snow accumulation score switches from a state equal to or greater than the second threshold to a state less than the second threshold.

[0109] However, in the snow accumulation detection system of this embodiment 1, the output delay time M can be set to an appropriate value of 0 seconds or more depending on the installation environment, making it possible to variably set the time from when it is determined that there is no snow accumulation to when the snow accumulation detection signal is switched to the OFF state.

[0110] Furthermore, the first threshold value used to determine whether there is snow in step S406 of FIG. 4 and step S1105 of FIG. 11 and the second threshold value used to determine whether there is no snow in step S1004 of FIG. 10 do not need to be the same value, and the second threshold value can be set to a value equal to or less than the first threshold value.

[0111] By setting the second threshold used to determine whether there is snow at a lower value than the first threshold used to determine whether there is snow at a lower value, it is possible to achieve an effect equivalent to setting the output delay time M. In addition, by setting the second threshold to a lower value than the first threshold, it is possible to provide a hysteresis characteristic to the switching between the determination whether there is snow at a higher value and the determination whether there is no snow at a higher value.

[0112] Next, the state transition of the snow accumulation score and the snow accumulation detection signal when a series of processes is executed, which is a combination of the snow accumulation detection process shown in Fig. 4 and the no-snow accumulation detection process shown in Fig. 10 and Fig. 11, will be summarized with reference to the drawings. Fig. 12 is a diagram showing an operation model of the snow accumulation detection system according to the first embodiment of the present disclosure.

[0113] Specifically, in Figure 12, the horizontal axis represents time and the vertical axis represents the snow accumulation score and snow melting device status, and this is an explanatory diagram showing the state transitions of the snow accumulation score and snow accumulation detection signal when a series of processes are executed by the control device 12.

[0114] Each of times T1 to T4 in FIG. 12 indicates a time when the control device 12 executes the following processes. T1: The time when the snow cover score first becomes equal to or greater than the first threshold and the number of judgments becomes n=1. T2: The time when the snow cover score is equal to or greater than the first threshold for N consecutive times. T3: The time when the snow cover score, which was equal to or greater than the second threshold, changes to less than the second threshold. T4: The time after the output delay time M seconds has elapsed since time T3

[0115] In FIG. 12, the first threshold value and the second threshold value are the same value, and are simply shown as threshold values.

[0116] By executing the snow detection process shown in FIG. 4 and the no-snow detection process shown in FIGS. 10 and 11, an operation model in which the snow score and snow detection signal transition as shown in FIG. 12 can be realized.

[0117] The technical features of the snow accumulation detection system according to the first embodiment can be summarized as follows. <Technical feature 1: Higher accuracy in detecting snow accumulation> The following first to fourth parameters can be adjusted to detect snow accumulation with high accuracy depending on the installation environment. 1st parameter: Target 1 shape and color settings By appropriately setting the first parameter according to the installation environment, it is possible to suppress deterioration in the detection position accuracy of the target 1 even when the camera 11 shakes due to the influence of weather.

[0118] In addition, by arranging the part where the road surface is exposed (corresponding to part A in Figure 3) of an appropriate size and position within the outline of target 1, the snow accumulation condition can be detected with high accuracy from changes in the correlation between the marking part and the road surface part.

[0119] Second parameter: First detection interval, consecutive count N By appropriately setting the second parameter according to the installation environment, it is possible to prevent erroneous detection of snow accumulation due to a parked vehicle, a user, or the like remaining on or passing over the target 1. In other words, by not turning the snow accumulation detection signal to the ON state unless a state in which snow accumulation is determined to exist continues for a certain period of time, it is possible to prevent the snow accumulation detection signal from changing from the OFF state to the ON state due to various disturbance factors.

[0120] -Third parameter: Sets the brightness range when performing binarization By appropriately setting the third parameter according to the installation environment, it is possible to obtain a binary image that distinguishes between snow, which is the intended detection target, and factors that may cause false detection. In other words, as explained with reference to Figure 6, by appropriately setting the brightness range for detecting snow according to the installation environment, it is possible to eliminate the influence of disturbance factors that have brightness outside the brightness range of snow.

[0121] - 4th parameter: Sets the first threshold used to detect the presence of snow By appropriately setting the fourth parameter according to the installation environment, the accuracy of detecting snow accumulation by comparison with the snow accumulation score can be improved.

[0122] As is also clear from Figure 12, by setting the first threshold higher, the ON period of the snow accumulation detection signal can be shortened, and conversely, by setting the first threshold lower, the ON period of the snow accumulation detection signal can be lengthened, and the first threshold can be appropriately set depending on the function of the external device driven by the snow accumulation detection signal.

[0123] <Technical feature 2: Higher accuracy in detecting the absence of snow> The following fifth to seventh parameters can be adjusted to enable highly accurate detection of snow-free conditions depending on the installation environment. 5th parameter: Target 1 shape and color setting Regarding the fifth parameter, it is possible to achieve the same effect as the first parameter when detecting the presence of snow.

[0124] - 6th parameter: Second detection interval setting By appropriately setting the fifth parameter according to the installation environment, the frequency of determining whether there is no snow can be adjusted. Note that the second detection interval according to the sixth parameter does not need to be set to the same value as the first detection interval included in the first parameter when detecting the presence of snow, and can be set individually according to the importance of switching the snow detection signal from the ON state to the OFF state.

[0125] - 7th parameter: Sets the second threshold used to detect the absence of snow. By appropriately setting the seventh parameter according to the installation environment, the accuracy of the detection of the absence of snow by comparison with the snow score can be improved. As described above, the second threshold value of the seventh parameter does not need to be set to the same value as the first threshold value, which is the fourth parameter, when detecting the presence of snow.

[0126] The snow accumulation detection system according to the first embodiment can be summarized as follows from the viewpoint of significant effects. (Effect 1) Improved snow accumulation detection level In order to detect the presence of snow with high accuracy based on a comparison of a reference image and a detected image through image processing, this disclosure uses a target 1 having a pattern that can be located through edge extraction processing and that can estimate the amount of snow through binarization processing.

[0127] Furthermore, in order to detect the presence of snow with high accuracy, the system is configured to allow the following four parameters to be pre-set to appropriate values ​​according to the installation environment: a first detection interval when acquiring detection images sequentially in chronological order, a brightness range for performing binarization processing, a first threshold value used in the comparison processing of snow accumulation scores, and a consecutive number that specifies the number of times the snow accumulation score will be equal to or greater than the first threshold value.

[0128] As a result, it is possible to detect the presence of snow with high accuracy by focusing on changes in the state of the target, rather than comparing images across the entire imaging range. Furthermore, by setting the four parameters to appropriate values, it is possible to prevent elements that are not caused by snow from being mistakenly detected as snow, even in environments with many obstacles to imaging. Therefore, it is possible to realize a snow detection system that can be adapted to various installation environments and can determine snow accumulation with improved detection accuracy.

[0129] (Effect 2) Improved detection of no snow cover When detecting a state without snow, a comparison process is performed between the reference image and the detected image using a target, just as when detecting a state with snow.

[0130] Furthermore, in order to detect a state without snow with high accuracy, the device is configured to allow the user to pre-set three parameters, namely, a second detection interval when acquiring detection images sequentially in chronological order, a second threshold value used in the comparison process of the snow accumulation scores, and an output delay time that dictates the elapsed time from when it is determined that the snow accumulation state has been lifted to when the snow accumulation detection signal is switched to the OFF state, to appropriate values ​​according to the installation environment.

[0131] As a result, it is possible to detect the absence of snow with high accuracy by focusing on the change in the state of the target, rather than comparing images across the entire imaging range. Furthermore, by setting the three parameters to appropriate values, it is possible to detect with high accuracy the decrease in snowfall and the change from a snowy state to a no-snow state, and to activate an external device such as a snow melting device at the desired timing. Therefore, it is possible to realize a snow detection system that can be adapted to various installation environments even in the absence of snow detection, and that can determine snow accumulation with improved detection accuracy. [Explanation of symbols]

[0132] 1 Target, 10 Snow accumulation detection system, 11 Camera, 12 Control device, 20 Snow melting device (external device).

Claims

1. A target provided in a monitoring range where snow accumulation state should be detected, for determining the snow accumulation state based on image processing, a camera that outputs an image of the target, and a control device that performs snow accumulation detection based on the image obtained from the camera are provided, the target has a pattern that can be located by edge extraction processing by the image processing and the snow accumulation amount on the target can be estimated by binarization processing by the image processing, the control device uses the image obtained from the camera when there is no snow accumulation as a reference image, at the time of inspection for determining the snow accumulation state, the images sequentially acquired from the camera at a preset first detection interval are used as detection images, for the reference image in the area of the target, an image after binarization processing with a preset luminance range is generated as a binarization reference image, for the detection image in the area of the target, an image after the binarization processing with the luminance range is generated as a binarization detection image, a difference image between the binarization reference image and the binarization detection image is calculated, and the number of pixels for which a difference exceeding a preset tolerance component is calculated for each pixel of the difference image is calculated as a snow accumulation score, and the snow accumulation score is sequentially calculated corresponding to the detection images sequentially acquired in the time series, when the state where the sequentially calculated snow accumulation scores are equal to or greater than a preset first threshold continues for a preset number of consecutive times, it is determined that it is a snow accumulation state, and a snow accumulation detection signal is output in an ON state Snow accumulation detection system.

2. The control device even after determining that it is the snow accumulation state, continues to perform the process of sequentially acquiring the detection images from the camera in the time series according to a preset second detection interval, generating the binarization detection image, and sequentially calculating the snow accumulation score, when the sequentially calculated snow accumulation score changes from a state where it is equal to or greater than a preset second threshold and less than or equal to the first threshold to a state where it is less than the second threshold, it is immediately determined that the snow accumulation state has been released, and the snow accumulation detection signal that was in the ON state is output in an OFF state The snow accumulation detection system according to Claim 1.

3. The control device outputs the snow accumulation detection signal in the OFF state after a preset output delay time has elapsed since determining that the snow accumulation state has been released The snow accumulation detection system according to claim 2.

Images