TW201743286A - Method for detecting of liquid - Google Patents

Abstract

The invention relates to a method for detecting of liquid which for a fluid level monitoring system. The fluid level monitoring system includes a light emitting module, a data capture module and a data processing module. Wherein, the light emitting module projects a planar laser beam to a wall and a liquid. And then, the common boundary of the wall and the liquid generates light segments which intersect at a point. The data capture module shooting the light segments to generate an original image. The data processing module strengthens and detects the original image for getting the image coordinate of the point at the light segments. Finally, transforms the image coordinate to a spatial coordinate which is the fluid level.

Description

Liquid level monitoring method

The invention relates to a liquid level monitoring method, in particular to projecting a planar laser beam onto a wall surface and a liquid surface, and at least three light segments forming a common intersection with the liquid surface and the liquid surface are processed by image processing. The method obtains a liquid level monitoring method for the spatial position of the co-intersection point.

Taiwan is the main path of typhoon invasion in the Pacific Northwest. It is a high-risk area and is highly vulnerable to natural disasters. In view of this, urban areas are prone to flooding due to the inability to eliminate excess rainfall or regional flooding in a short period of time due to the large amount of rainwater, which seriously threatens the lives and property of the people. Accordingly, liquid level monitoring and early warning of river water level has always been the primary target of disaster prevention; and in order to achieve the above objectives, it is preferable to use a liquid level monitoring method to monitor the liquid level position. In addition, the liquid level monitoring method can be widely applied to various other fields such as chemical or medical related fields. For example, in experiments in the field of chemistry, the measurement of various liquid chemical materials is very important because of the different concentration configurations between various chemical materials. Furthermore, in the medical related field, for monitoring the dosage of the medicament used by the patient, for example, when the drip is administered, whether the medicine of the drip bottle has fallen below a critical value, and the medical personnel need to be reminded to supplement the medicament or stop applying. Drip.

The liquid level monitoring method may be a conventional liquid level measuring method for measuring the liquid surface position of a liquid by taking a surface image of a liquid, and the embodiment may be referred to as the Republic of China No. 201024687 "Laser". Patent of optical image water level measuring device and method thereof. on The method of the patent case uses the light emitted by the two laser light sources to form two laser spots on the surface of a water body. Then, an image capturing device is used to capture and obtain a water body image including the two laser light spots, and the water body image is transmitted to the image processing device for analysis. The analysis method of the image processing apparatus calculates the distance between the two laser spots using a corrected regression curve, thereby obtaining the water level height of the water body.

However, the two laser light spots are generated by the two laser light sources, and subsequent operations are performed to determine the water level height of the water body. In view of the above, in order to simplify the hardware requirements of the above liquid level measuring method, the present invention provides a liquid level monitoring method, which only needs to project a planar laser beam onto a wall surface and a liquid surface, the wall surface and the liquid At least three ray segments of a common intersection are formed at the surface intersection, and the position of the co-intersection is analyzed to accurately measure and obtain the liquid level position.

The object of the present invention is to provide a liquid level monitoring method, in which a planar laser beam is projected by a light-emitting module onto a wall surface and a liquid surface, and at least three light segments are formed at a boundary point between the wall surface and the liquid surface. And analyze the position of the co-intersection point to accurately measure and obtain the liquid level position.

A liquid level monitoring method is applied to a liquid level monitoring system for monitoring a liquid level position of a liquid in a liquid holding device, the liquid level monitoring system comprising a data processing module, the method comprising: using the data processing module Reading an original image; detecting, by the data processing module, at least three light segments forming a common intersection in the original image; calculating, by the data processing module, image coordinates of the common intersection; and using the data processing module to The image coordinates are converted into a space coordinate, which is the liquid level of the liquid.

Wherein, the liquid receiving device has a wall surface in contact with the liquid, so that the liquid level of the liquid is in contact with the wall surface to form a line of intersection, the liquid level monitoring system further comprises a light emitting module and a data capturing module. The data processing module is coupled to the light emitting module and the data capturing module. The method projects a plane toward the intersection line before the data processing module reads the original image. a laser beam for forming the at least three light segments adjacent to the wall surface and the liquid surface of the intersection line, wherein the intersection point is located on the intersection line, and the data capture module is photographed toward the at least three light segments The original image is produced.

The planar light beam forms a first light segment and a second light segment on the wall surface and the liquid surface, and the second light segment forms a third light segment on the wall surface, and the third light segment is The liquid surface forms a fourth light segment, and the at least three light segments include at least three light segments of the first light segment, the second light segment, the third light segment, and the fourth light segment.

The method for detecting the at least three ray segments performs an edge detection and a morphological processing on the original image to generate a binary image, and performs line segment detection on the binary image.

Wherein, the morphological processing performs filling and skeletonization in the morphology of the original image.

The data processing module performs line detection and curve detection on the original image when the at least three light segments comprise at least one curved line segment.

Wherein, when the wall surface is a curved wall surface, the at least three light segments comprise at least one curved line segment, and the data processing module performs line detection and curve detection on the original image.

The method for calculating the image coordinates of the co-intersection point by the data processing module is a least squares method.

The method for converting the image coordinates into the space coordinates by the data processing module is a direct linear conversion method.

1‧‧‧ Data Processing Module

2‧‧‧Lighting module

3‧‧‧Data Capture Module

4‧‧‧Database module

S1‧‧‧ pre-processing program

S11‧‧‧ parameter calibration steps

S12‧‧‧Image reading step

S13‧‧‧ Grayscale processing steps

S2‧‧‧ Feature Enhancement Procedure

S21‧‧‧ Edge Detection Steps

S22‧‧‧ Morphological processing steps

S3‧‧‧ image analysis program

S31‧‧‧ Line segment detection steps

S32‧‧‧Complex calculation steps

S33‧‧‧ Liquid level estimation procedure

L‧‧‧Liquid

S‧‧‧ liquid level

W‧‧‧ wall

A‧‧‧ original image

T‧‧‧ planar laser beam

U‧‧‧First light segment

Y‧‧‧second light segment

Z‧‧‧The third light segment

X‧‧‧fourth light segment

G‧‧‧ image gradient values

E‧‧‧ image boundary value

ω ‧‧‧ threshold

Fig. 1 is a schematic view showing the system architecture of the embodiment of the liquid level monitoring method of the present invention.

Fig. 2 is a flow chart showing the hardware operation of the embodiment of the liquid level monitoring method of the present invention.

Figure 3 is a flow chart showing the software operation of the embodiment of the liquid level monitoring method of the present invention.

Figure 4 is a schematic diagram of light segment detection in the embodiment of the liquid level monitoring method of the present invention.

The above and other objects, features and advantages of the present invention will become more <RTIgt; "Coupled Connection" means that two electronic devices can communicate with each other via wired or wireless technology, but not limited thereto, as will be understood by those of ordinary skill in the art to which the present invention pertains.

"Pixels" as used throughout the present invention refers to a minimum unit of image composition for indicating the resolution of the image. For example, if the resolution of the image is 1024×768, it means The image has a total of 1024 x 768 pixels, which is understood by those of ordinary skill in the art to which the present invention pertains.

The "Color Level" as used throughout the present invention refers to the degree of gradation of the color component or brightness exhibited by the pixel, for example, the red (R), green (G), and blue (B) components of the color image. The range of the gradation ranges from 0 to 255; alternatively, the range of the gradation of the grayscale image (Luminance) may range from 0 to 255, which can be understood by those of ordinary skill in the art to which the present invention pertains.

Please refer to FIG. 1 , which is a schematic diagram of a system architecture of an embodiment of a liquid level monitoring method of the present invention, comprising: a data processing module 1 , a light emitting module 2 , and a data capturing module 3 . The light module 2 and the data capture module 3 are coupled to the data processing module 1 . The system can be used to monitor the level of the liquid surface S of a liquid L in a liquid holding device (for example, a water channel or a container). The liquid receiving device comprises two side walls, a bottom wall and an accommodating space. The accommodating space is formed by the two side walls and the bottom wall, and the accommodating space is for accommodating the liquid L. In detail, one of the two side walls has a wall surface W facing the accommodating space, and the wall surface W can be a planar wall surface or a curved wall surface, and the wall surface W has an opposite top end and a bottom end. Further, in the case where the liquid L is contained in the liquid holding device, the liquid L is in contact with the wall surface W, And the liquid surface S is in contact with the wall surface W to form an intersection line. In addition, the data processing module 1 is preferably coupled to a database module 4, such as a database of MySql, Oracle, etc., and the database module 4 can be used to store the original image captured by the data capture module 3. A, the data processing module 1 can also receive an image as the original image A and use it as a subsequent image processing analysis.

Please refer to FIG. 2, which is a hardware operation flowchart of the embodiment of the liquid level monitoring method of the present invention. The light-emitting module 2 projects a planar laser beam T toward the intersection line, and the planar laser beam T preferably covers the connection direction between the light-emitting module 2 and the bottom end of the wall surface W. The range in which the light-emitting module 2 and the tip end of the wall surface W are connected. Also, the planar laser beam T is not parallel to the intersection. Thereby, in the case where the liquid L is contained in the liquid containing device, the planar laser beam T can form at least three light segments with the liquid surface S adjacent to the wall surface W of the intersection line, and the at least three light segments are A common intersection is formed on the intersection. In this embodiment, the planar laser beam T forms a first ray segment U and a second ray segment Y on the wall surface W and the liquid surface S, respectively, and the second ray segment Y forms a first layer on the wall surface W. a third ray segment Z, the third ray segment Z forming a fourth ray segment X on the liquid surface S, the at least three ray segments including the first ray segment U, the second ray segment Y, and the third ray segment Z And at least three ray segments in the fourth ray segment X. In addition, the wall surface W is described in the following manner as a wall surface of a channel side wall. Furthermore, the illumination module 2 can be a Laser Emitter with Prism Lens, and a laser beam can be used to project a beam of light onto a stack, thereby forming the planar laser beam T. . However, the embodiment of the light-emitting module 2 and the container to which the monitoring method is applied are not limited to the above-described types.

Please refer to FIG. 2 again. The data capture module 3 (for example, a surveillance camera, a network camera or an infrared camera) can be coupled to the data processing module 1 (for example, a computer host, a file server, or a cloud). The server, etc.) acts as the system execution architecture. In this embodiment, the data capture module 3 can be disposed adjacent to the light-emitting module 2, and capture an original image A (Original Image) including the at least three light segments, for example, single or continuous. (Continued) image or the like, the original image A may be a color or grayscale image. The data processing module 1 can receive the original image A from the data acquisition module 3, and execute the software operation process disclosed in the embodiment of the liquid level monitoring method of the present invention to measure the liquid surface position, but This is limited to this. In detail, the above liquid level position is the water level height of the liquid L. In addition, the original image A is followed by a color image as an implementation aspect, but not limited thereto, and the like, and can be applied to liquid level position measurement of black and white or continuous images, which is a technical field to which the present invention pertains. Those who have the usual knowledge can understand it, and I won't go into details here.

Referring to FIG. 3, it is a software operation flowchart of the embodiment of the liquid level monitoring method of the present invention, which may include a pre-processing program S1, a feature enhancement program S2, and an image analysis program S3, respectively, as described later.

The pre-processing program S1 may include a parameter calibration step S11, an image reading step S12, and a grayscale processing step S13. The parameter calibration step S11 can be configured by the data processing module 1 to set at least four external control points to obtain external parameters of the data capture module 3. The data processing module 1 calculates the internal parameters of the data capture module 3 by calculating the external parameters of the data capture module 3, a plane correction plate, and a correction formula. In this embodiment, the data capture module 3 is a camera. The internal parameters of the camera can be used to correct the Radiation Distortion of the lens, which can be understood by those of ordinary skill in the art to which the present invention pertains, and will not be further described herein. Among them, the correction formula can refer to "a correction method for solving image distortion provided by Holland et al. (1997)". In addition, the calculation method of the internal parameters of the data capture module 3 is understood by those having ordinary knowledge in the technical field of the present invention, and details are not described herein.

The image reading step S12 can read the original image A from the data capturing module 3 by the data processing module 1 , but not limited thereto. The original image A may include an image of the wall surface W, the liquid L, and the at least three light segments.

The grayscale processing step S13 can be performed by the data processing module 1 on the original image A. The gray scale processing is performed according to the gradation of the red, green and blue components of each pixel of the original image A, and the gradation of the original image A is averaged to a luminance ranging from 0 to 255.

However, the parameter calibration step S11 and the gray scale processing step S13 can be selectively performed, for example, the internal and external parameters of the data capture module 3 need not be calibrated, and the parameter calibration step S11 can be omitted; Furthermore, the color gradation of the pixel of the original image A is in the range of 0 to 255, and the grayscale processing step S13 can be omitted, which is understood by those skilled in the art to which the present invention pertains, and is not limited thereto.

The main purpose of the feature enhancement program S2 is to enhance the features of the light segments included in the original image A, and may include an edge detection step S21 and a morphological processing step S22. The edge detection step S21 can detect the edge feature of the original image A by the data processing module 1. The detection method can be a gradient operation sub-edge search method, for example, Canny Edge Detection. Or Sobel Edge Detection, etc. In this embodiment, the Kenny edge detection is used as the implementation description as a follow-up description, but not limited thereto, the main principle is to calculate the image gradient (Gradient) value G of each pixel in the original image A, and according to The image gradient value G calculates the image boundary value E of each pixel. In detail, the original image A can be a binary image in the image processing after the edge detection step S21 is performed, but is not limited thereto. The image boundary value E is 1 or 0, which represents that one pixel is a boundary or not a boundary. The calculation formula of the image gradient value G and the image boundary value E is as shown in the following equations (1) to (3):

Wherein, f (x, y) based upon the grayscale values of the pixels of the original image A that is the (x, y), and 0 x<M and 0 y<N; G _x represents the gradient value of the horizontal direction of the original image A; G _y represents the gradient value of the vertical direction of the original image A; ω is a threshold value. If the image gradient value G of the pixel ( x , y ) of the original pixel A is greater than the threshold ω , the image boundary value E ( x , y ) of the pixel ( x , y ) is set to 1; if the original pixel A When the image gradient value G of the pixel ( x , y ) is not larger than the threshold value ω , the image boundary value E ( x , y ) of the pixel ( x , y ) is set to be 0. In this embodiment, the data processing module 1 detects the edge features of the original image A by using the gradient operation sub-edge search method to obtain the wall surface and the liquid surface of the original image A whose boundary value is 1. The at least three light segments.

The morphological processing step S22 can use the morphological image processing method of the fill and skeletoning of the original image A by the data processing module 1 to generate the at least three ray segments in the original image A. feature. The original image A can be a binary image in the image processing after the morphological processing step S22 is performed, but is not limited thereto. For example, after the data processing module 1 uses the filled morphological image processing method for the original image A, the edge features of at least three ray segments included in the original image A are filled inside. When the at least three ray segments are all in a straight line, the at least three ray segments are each a mxn quadrilateral, wherein m represents a pixel value of the quadrilateral width, and m ≧ 1; n represents a pixel value of the quadrilateral length, and N≧1; when the curve is included in the at least three ray segments, the pixel value of the width of the curve is m, and m ≧ 1. However, when the pixel value of the width or length of the at least three light segments is greater than 1, the subsequent image processing steps may be misjudged. Therefore, the data processing module 1 uses the skeletonized morphological image processing on the original image A. The method converts pixel values of one of the widths or lengths of the at least three light segments into 1 pixel. Thereby, the accuracy of the subsequent image processing steps and the image processing efficiency are improved.

Referring to FIG. 4 together, the image analysis program S3 may include a line segment detecting step S31, a common point calculating step S32, and a liquid level estimating step S33. The line segment detecting step S31 can detect at least three light segments included in the original image A by the data processing module 1. In this embodiment, when the wall surface W is the planar wall surface, a Hough Transform method can be used to detect "Hough (1962)", and at least three rays included in the original image A are detected. The main principle of the segment is to convert the x and y coordinates of each point of each line segment in the original image A into ρ(rho) and θ(theta) polar coordinates. The conversion formula is as shown in the following equations (4) to (5):

According to the Hough transform method, the line segments in the same direction in the original image A have the characteristics of the intersection at the same point on the ρ and θ polar graphs. Therefore, the point of the maximum intersection amount can be taken out, and the optimum values of ρ and θ are selected. The line segment is the detected light segment.

However, when the wall surface W is the curved wall surface, the at least three light segments may include at least one curved line segment. Therefore, the data processing module 1 may perform line detection and curve detection on the original image A.

The common point calculation step S32 can obtain the line segment equation of the at least three light segments from the data processing module 1. In this embodiment, the data processing module 1 can take any pixel coordinate points that constitute the minimum required number of the at least three light segments in the original image A (for example, at least two pixel coordinate points are required to generate a linear equation or at least A three-pixel coordinate point is required to generate a curve equation) to generate a line segment equation for the at least three light segments. In addition, when the at least three light segments are all straight line segments, the data processing module 1 can represent the at least three light segments in a vector representation to represent the respective line segment equations. Then, the data processing module 1 can generate a matrix equation according to the line segment equation described above, and use the data processing module 1 to formulate the matrix equation by a joint legislation or a least square method (Generalized Least Squares). The calculation is performed to obtain image coordinates of the co-intersection points of the at least three light segments. In this embodiment, the data processing module 1 uses the least squares method to calculate the line segment equation, but not limited thereto, and the operation method of the least square method has the usual knowledge in the related technical field to which the present invention pertains. It can be understood that there will be no more details here.

For example, when the wall surface is the planar wall surface, the data processing module 1 can detect two straight line segments L1, L2 in the original image A. The image coordinates of any two points of the straight line segment L1 are (248, 216) and (276, 538), respectively; and the image coordinates of any two points of the straight line segment L2 are (260, 387) and (523, 825), respectively. The vector representation of a straight line segment is as shown in the following equation (6): a _i +tn _i ,- <t< (6) where a _i is the matrix representation of the pixel coordinates of the i-th straight line segment, and n _i is the unit vector of the i-th straight line segment, and the unit vectors of the two straight line segments L1 and L2 are respectively [0.0866 0.9962 ] ^T and [0.5148 0.8573] ^T . Then, the data processing module 1 calculates the image coordinates of the co-intersection points of the two straight line segments L1 and L2 by the least square method, wherein the least squares method is calculated as shown in the following equations (7) to (9). :p=R ^-1 q (7)

Where p is the coordinate of the co-intersection point, j(j=1, 2,..., J) is the j-th straight line segment, I is the unit matrix, and n _j is the unit vector of the j-th straight line segment, For the transfer matrix of the unit vector of the j-th straight line segment, a _j is the vector representation of the j-th straight line segment. And the image coordinates of the co-intersection point are obtained (263, 392).

The liquid level estimating step S33 can be converted by the data processing module 1 into a spatial coordinate of the wall surface and the liquid surface boundary, thereby estimating the liquid level position. In this embodiment, the conversion of the space coordinates preferably adopts a direct linear transformation method, and the formula is as shown in the following formula (10): Where (u, v) is the image coordinate of the co-intersection point, and (X, Y, Z) is the space coordinate of the boundary between the wall surface and the liquid surface, L _i (i=1, 2, ..., 11) The external parameters of module 3 are retrieved for this data. Since the data capture module 3 captures the wall surface W of the channel side wall as an implementation aspect, it is assumed that the distance between the data capture module 3 and the channel is such that the data capture module 3 does not need to consider the depth of field. influences. Therefore, the data processing module 1 can set the Y coordinate of the space coordinate to be zero, so as to reduce the time required for the liquid level estimation step S33, which has the effect of improving the "image processing efficiency". Therefore, the formula of the direct linear conversion method can be modified as shown in the following formula (11): Where (u, v) is the coordinate point of the co-intersection point, and (X, Y, Z) is the space coordinate of the boundary between the wall surface and the liquid surface, L _i (i=1, 3, 4, 5, 7, 8 , 9, 11) Extract the external parameters of the module 3 for the data.

In summary, the liquid level monitoring method of the present invention sets the four external control points to the data processing module 1 in the image pre-processing program S1 to obtain the external parameters of the data capturing module 3. The data processing module 1 generates internal parameters of the data capture module 3 by using the external parameter, the plane calibration plate, and the correction formula. Then, the data processing module 1 reads the original image A from the data capturing module 3, and performs image processing on the original image A by executing the feature enhancement program S2. Furthermore, the data processing module 1 executes the image analysis program S3 to detect features of at least three light segments included in the original image A, and calculates image coordinates of the intersection points of the at least three light segments to estimate the wall surface. The space coordinates at the junction with the liquid level, and the liquid level position is obtained. Accordingly, the liquid level monitoring measurement method of the present invention can achieve the purpose of accurately measuring the liquid level position.

While the invention has been described in connection with the preferred embodiments described above, it is not intended to limit the scope of the invention. The technical scope of the invention is protected, and therefore the scope of the invention is defined by the scope of the appended claims.

S1‧‧‧ pre-processing program

S11‧‧‧ parameter calibration steps

S12‧‧‧Image reading step

S13‧‧‧ Grayscale processing steps

S2‧‧‧ Feature Enhancement Procedure

S21‧‧‧ Edge Detection Steps

S22‧‧‧ Morphological processing steps

S3‧‧‧ image analysis program

S31‧‧‧ Line segment detection steps

S32‧‧‧Complex calculation steps

S33‧‧‧ Liquid level estimation procedure

Claims (9)

A liquid level monitoring method is applied to a liquid level monitoring system for monitoring a liquid level position of a liquid in a liquid holding device, the liquid level monitoring system comprising a data processing module, the method comprising: using the data processing module Reading an original image; detecting, by the data processing module, at least three light segments forming a common intersection in the original image; calculating, by the data processing module, image coordinates of the common intersection; and using the data processing module to The image coordinates are converted into a space coordinate, which is the liquid level of the liquid. The liquid level monitoring method according to claim 1, wherein the liquid receiving device has a wall surface in contact with the liquid, so that the liquid level of the liquid is in contact with the wall surface to form a line of intersection, the liquid level monitoring The system further includes a light-emitting module and a data capture module, the data processing module is coupled to the light-emitting module and the data capture module, wherein the data processing module reads the original image before The illuminating module projects a planar laser beam toward the intersection line to form the at least three ray segments adjacent to the wall surface of the intersection line, and the co-intersection point is located on the intersection line, and the data is used The module captures the at least three light segments and produces the original image. The liquid level monitoring method of claim 2, wherein the planar laser beam forms a first light segment and a second light segment on the wall surface and the liquid surface, and the second light segment is The third light segment forms a fourth light segment on the liquid surface, and the at least three light segments include the first light segment, the second light segment, the third light segment, and the At least three ray segments in the fourth ray segment. The liquid level monitoring method of claim 2, wherein, when the wall surface is a curved wall surface, the at least three light segments comprise at least one curved line segment, and the data processing module performs line detection on the original image. And curve detection. The liquid level monitoring method of claim 1, wherein the data processing module performs line detection on the original image when the at least three light segments comprise at least one curved line segment Curve detection. The liquid level monitoring method of claim 1, wherein the detecting method of the at least three light segments performs an edge detection and a morphological processing on the original image to generate a binary image, and Perform line segment detection on the binary image. The liquid level monitoring method according to claim 6, wherein the morphological processing performs filling and skeletonization in the morphology of the original image. The liquid level monitoring method of claim 1, wherein the data processing module calculates the image coordinates of the co-intersection point as a least squares method. The liquid level monitoring method according to claim 1, wherein the data processing module converts the image coordinates into the space coordinates by a direct linear conversion method.