CN103148881B - Method and device for automatically reading pointer instrument by video frequency - Google Patents

Abstract

The invention discloses a method and a device for automatically reading a pointer instrument by video frequency. The method comprises the following steps of: obtaining an initial scale point and a stop scale point from an effective region of an input image calibration dial plate, and recording an initial position scale value and a stop position scale value; detecting all line segments within the effective region of the dial plate; setting rule parameters, selecting the detected line segments in a regularization way, and filtering out a meter needle; and reading the scale appointed by the meter needle by a corresponding algorithm according to the type of the dial plate, and obtaining the reading number. After the invention is adopted, a plurality of needle targets can be quickly and exactly detected, any pre-estimation treatments (such as colors, positions and the like of the needle dial plate) are not needed, and high in accuracy rate.

Description

Method and device for automatically reading pointer type instrument by using video

Technical Field

The invention relates to a dial plate detection technology, in particular to a method and a device for automatically reading a pointer instrument by using video.

Background

In a substation of a power system, meters are commonly used and play an important role in ensuring safe operation of the system, wherein a large number of pointer meters are not available, and remote acquisition of readings of the pointer meters has not been a comprehensive method.

The prior art mainly comprises an instrument monitoring method based on an object outline, a pointer instrument reading method based on a scale inquiry table and the like. The basic scheme of the former method is that a master control end computer controls an inspection robot in a transformer substation through a wireless local area network based on a remote meter detection/reading module which stores a circle/ellipse/rectangle search and detection algorithm based on an object outline, locks the spatial position of an instrument through a search subprogram in a set search space range according to a preset inspection route, then obtains a geometric image of the instrument by using a visible light camera, and sends the geometric image back to the master control end through the wireless local area network for data processing so as to achieve the purpose of remote monitoring. The method needs to accurately control the accurate positioning of the inspection robot so as to ensure the accuracy of image acquisition, and has no robustness on dial plate searching. In the latter method, a database is formed by establishing scale lookup tables of different instruments; collecting an instrument image, inputting the instrument image and preprocessing the instrument image so as to extract an effective identification area of an instrument dial; extracting the pointer and the scale mark of the instrument from the effective identification area of the instrument dial by using the stroke characteristics of the pointer and the scale mark; and matching the extracted scale marks with corresponding scale query tables in a scale query table database according to the extracted pointer and the extracted scale marks, and automatically identifying the reading. Compared with the former method, the method is improved to a certain extent, but the method can only be suitable for fixed monitoring points, the dial plate searching is not robust, particularly a complex database needs to be established, and the workload at the early stage is very large.

Disclosure of Invention

Therefore, it is necessary to provide a method and an apparatus for automatically reading a pointer instrument by using video, which are capable of adapting to a reading method of a pointer instrument with better adaptability, and are insensitive to environmental changes, more robust, and simpler in preprocessing at an early stage.

A method relating to the automatic reading of a pointer instrument using video, comprising:

acquiring a starting scale point and an ending scale point from an effective area of an input image calibration dial, and recording a scale value of a starting position and a scale value of an ending position;

detecting all line segments in the effective area of the dial plate;

setting rule parameters, carrying out regularized selection on the detected line segments, and filtering out an instrument pointer;

and reading the scale pointed by the instrument pointer by adopting a corresponding algorithm according to the type of the dial plate to obtain the reading.

Accordingly, an apparatus relating to automatic reading of a pointer instrument using video, comprising:

the image marking unit is used for marking the effective area of the dial plate from the input image, acquiring a starting scale point and an ending scale point, and recording a scale value of the starting position and a scale value of the ending position;

the line segment detection unit is connected with the image marking unit and is used for detecting all line segments in the effective area of the dial plate;

the pointer filtering unit is connected with the line segment detecting unit and used for setting rule parameters, performing regularized selection on the detected line segments and filtering out the meter pointers;

and the operation reading unit is connected with the pointer filtering unit and used for reading the scale pointed by the instrument pointer by adopting a corresponding algorithm according to the type of a dial plate to obtain a reading.

The implementation of the invention has the following beneficial effects:

the invention can quickly and accurately detect a plurality of pointer targets only by roughly calibrating the dial area without any pre-estimation processing (such as the color, the position and the like of the pointer dial), and has high accuracy. According to the given calibration method, universal dial scale detection can be formed, and an instrument scale database does not need to be established. And the method is insensitive to environmental changes and has better robustness. The internal parameter setting is less, and the commonality is stronger.

Drawings

FIG. 1 is a flow chart of a method for automatically reading a pointer instrument using video according to the present invention;

FIG. 2 is a flow chart of an embodiment of the invention for disc identification using a method for automatically reading a pointer instrument using video;

FIG. 3 is a flow chart of an embodiment of square plate recognition of the method for automatically reading a pointer instrument using video according to the present invention;

FIG. 4 is a schematic diagram of an apparatus for automatically reading a pointer instrument using video according to the present invention;

fig. 5 is a schematic diagram of an embodiment of the device for automatically reading a pointer instrument by using video.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings.

Fig. 1 is a flow chart of a method for automatically reading a pointer instrument by using video according to the present invention, which includes:

s101: acquiring a starting scale point and an ending scale point from an effective area of an input image calibration dial, and recording a scale value of a starting position and a scale value of an ending position;

s102: detecting all line segments in the effective area of the dial plate;

s103: setting rule parameters, carrying out regularized selection on the detected line segments, and filtering out an instrument pointer;

s104: and reading the scale pointed by the instrument pointer by adopting a corresponding algorithm according to the type of the dial plate to obtain the reading.

The invention provides a simpler, more intelligent and automatic scale marking algorithm, and an effective dial area can be identified and marked by the algorithm provided by the method only through simple initial marking. The method can automatically adapt to the physical profiles of various dials and adopt corresponding algorithms to identify the current position of the pointer and calculate corresponding readings without establishing a meter scale database in advance.

The method can quickly and accurately detect a plurality of pointer targets only by roughly calibrating the dial area, does not need any pre-estimation processing (such as the color, the position and the like of the pointer dial), and has high accuracy. According to the given calibration method, universal dial scale detection can be formed, and an instrument scale database does not need to be established. And the method is insensitive to environmental changes and has better robustness. The internal parameter setting is less, and the commonality is stronger.

In one embodiment, the step of detecting all the line segments in the active area of the dial comprises:

performing Gaussian smoothing on the effective region to obtain a smooth image;

calculating the gradient and the edge line angle of the smooth image;

establishing a search list, and putting the position coordinates of each point into a corresponding BIN according to the gradient amplitude of each point in a descending order;

creating an auxiliary image for temporarily recording the identification information, setting the gradient amplitude smaller than theta as used, and setting the other as unused, wherein,

q is the pixel quantization precision, τ is the edge similarity angle precision;

the search is started by the pixel in the search list, starting with the largest gradient value in BIN and marked as unused.

1. And performing Gaussian smoothing on the dial area to obtain a smooth image.

2. The gradient and edge line angle of the smoothed image are calculated.

$g_x=\frac{I(x+1,y+1)+I(x+1,y)-I(x,y)-I(x,y+1)}{2}---(2-1)$

$g_y=\frac{I(x+1,y+1)+I(x,y+1)-I(x,y)-I(x+1,y)}{2}---(2-2)$

$G=\sqrt[2]{g_x^2(x,y)+g_y^2(x,y)}---(2-3)$

$M_A=\arctan \left(\frac{g_x(x,y)}{-g_y(x,y)}\right)---(2-4)$

3. And establishing a search list, and putting the position coordinates of each point into the corresponding BIN according to the gradient magnitude of each point in the order of magnitude.

4. An auxiliary image (template image, temporarily used for recording identification information) is created, and the image with gradient amplitude smaller than theta is set as used, and the other image is set as unused.

<math> <mrow> <mi>&theta;</mi> <mo>=</mo> <mfrac> <mi>q</mi> <mrow> <mi>sin</mi> <mi>&tau;</mi> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>-</mo> <mn>5</mn> <mo>)</mo> </mrow> </mrow> </math>

Where q is the pixel quantization precision and τ represents the edge similarity angle precision.

5. Starting the search by pixels in the search list, the search rule: starting with the largest gradient value in BIN and marked as unused. In particular, the amount of the solvent to be used,

firstly, setting a pixel P as a seed pixel, and growing a region;

growth rule:

(1) have similar edge line angles, i.e., at edge angles [ Pma- τ, Pma + τ ];

(2) k neighborhood growth;

(3) and setting the pixel which is successfully grown as used.

Secondly, calculating the center and the main direction of the grown rectangular region

The center of the rectangle is as follows:

<math> <mrow> <msub> <mi>c</mi> <mi>x</mi> </msub> <mo>=</mo> <mfrac> <mrow> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <mi>G</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> <mo>*</mo> <mi>x</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> </mrow> <mrow> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <mi>G</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>-</mo> <mn>6</mn> <mo>)</mo> </mrow> </mrow> </math>

<math> <mrow> <msub> <mi>c</mi> <mi>y</mi> </msub> <mo>=</mo> <mfrac> <mrow> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <mi>G</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> <mo>*</mo> <mi>y</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> </mrow> <mrow> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <mi>G</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>-</mo> <mn>7</mn> <mo>)</mo> </mrow> </mrow> </math>

where j varies within a rectangular area.

Main direction of the rectangle:

constructing feature matrices $\left[\begin{array}{cc}{m}^{\mathrm{xx}}& m^{\mathrm{xy}}\\ m^{\mathrm{yx}}& m^{\mathrm{yy}}\end{array}\right]---(2-8)$

<math> <mrow> <msup> <mi>m</mi> <mi>xx</mi> </msup> <mo>=</mo> <mfrac> <mrow> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <mi>G</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> <mo>*</mo> <msup> <mrow> <mo>(</mo> <mi>x</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> <mo>-</mo> <msub> <mi>c</mi> <mi>x</mi> </msub> <mo>)</mo> </mrow> <mn>2</mn> </msup> </mrow> <mrow> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <mi>G</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>-</mo> <mn>9</mn> <mo>)</mo> </mrow> </mrow> </math>

<math> <mrow> <msup> <mi>m</mi> <mi>xy</mi> </msup> <mo>=</mo> <msup> <mi>m</mi> <mi>yx</mi> </msup> <mo>=</mo> <mfrac> <mrow> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <mi>G</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> <mo>*</mo> <mrow> <mo>(</mo> <mi>x</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> <mo>-</mo> <msub> <mi>c</mi> <mi>x</mi> </msub> <mo>)</mo> </mrow> <mo>*</mo> <mrow> <mo>(</mo> <mi>y</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> <mo>-</mo> <msub> <mi>c</mi> <mi>y</mi> </msub> <mo>)</mo> </mrow> </mrow> <mrow> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <mi>G</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>-</mo> <mn>10</mn> <mo>)</mo> </mrow> </mrow> </math>

<math> <mrow> <msup> <mi>m</mi> <mi>yy</mi> </msup> <mo>=</mo> <mfrac> <mrow> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <mi>G</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> <mo>*</mo> <msup> <mrow> <mo>(</mo> <mi>y</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> <mo>-</mo> <msub> <mi>c</mi> <mi>y</mi> </msub> <mo>)</mo> </mrow> <mn>2</mn> </msup> </mrow> <mrow> <msub> <mi>&Sigma;</mi> <mi>j</mi> </msub> <mi>G</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>-</mo> <mn>11</mn> <mo>)</mo> </mrow> </mrow> </math>

Calculating its feature vector as its main direction of the rectangle

And thirdly, calculating the density of the edge angles of the rectangular block, if the density is larger than a threshold value D representing a threshold parameter, trimming the rectangular area, and repeatedly calculating the density of the edge angles.

Density calculation formula:

$d=\frac{k}{\mathrm{area}\left(\mathrm{rect}\right)}---(2-12)$

where k denotes the number of edge angles in the rectangular region equal to the main direction of the rectangular region, and area (rect) denotes the area of the rectangular region.

Fourthly, calculating the false alarm rate of the edge angle in the rectangular block as a verification formula

<math> <mrow> <mi>NFA</mi> <mrow> <mo>(</mo> <mi>rect</mi> <mo>)</mo> </mrow> <mo>=</mo> <msup> <mrow> <mo>(</mo> <mi>NM</mi> <mo>)</mo> </mrow> <mrow> <mn>5</mn> <mo>/</mo> <mn>2</mn> </mrow> </msup> <mo>*</mo> <mi>&gamma;</mi> <mo>*</mo> <msubsup> <mi>&Sigma;</mi> <mrow> <mi>j</mi> <mo>=</mo> <mi>k</mi> </mrow> <mi>n</mi> </msubsup> <mfenced open='(' close=')'> <mtable> <mtr> <mtd> <mi>n</mi> </mtd> </mtr> <mtr> <mtd> <mi>j</mi> </mtd> </mtr> </mtable> </mfenced> <msup> <mi>p</mi> <mi>j</mi> </msup> <msup> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>p</mi> <mo>)</mo> </mrow> <mrow> <mi>n</mi> <mo>-</mo> <mi>j</mi> </mrow> </msup> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>-</mo> <mn>13</mn> <mo>)</mo> </mrow> </mrow> </math>

Wherein,

<math> <mrow> <mfenced open='(' close=')'> <mtable> <mtr> <mtd> <mi>n</mi> </mtd> </mtr> <mtr> <mtd> <mi>j</mi> </mtd> </mtr> </mtable> </mfenced> <mo>=</mo> <mfrac> <mrow> <mi>&Gamma;</mi> <mrow> <mo>(</mo> <mi>n</mi> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> <mrow> <mi>&Gamma;</mi> <mrow> <mo>(</mo> <mi>j</mi> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>*</mo> <mi>&Gamma;</mi> <mrow> <mo>(</mo> <mi>n</mi> <mo>-</mo> <mi>j</mi> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>-</mo> <mn>14</mn> <mo>)</mo> </mrow> </mrow> </math>

m, N denotes the height and width of the image, and γ denotes the statistical number of edge angles different from the main direction in the rectangular region.

And fifthly, if the false alarm rate is greater than the threshold value e, trimming the rectangular area, otherwise, adding the rectangular line segment into an output queue.

FIG. 2 is a flow chart of an embodiment of the invention for disc identification in a method for automatically reading a pointer instrument using video.

S201: acquiring a starting scale point and an ending scale point from an effective area of an input image calibration dial, and recording a scale value of a starting position and a scale value of an ending position;

s202: calculating a perpendicular bisector of the start tick mark and the end tick mark;

s203: and calculating the position of the center of the circle on the middle vertical line by correcting the position of the point and the scale value.

S204: detecting all line segments in the effective area of the dial plate;

s205: setting rule parameters, carrying out regularized selection on the detected line segments, and filtering out an instrument pointer;

s206: for the pointer line segment after the regularization, when the slope angle kp 1-kp 2 belongs to [ -eps, + eps ], then the two straight lines are merged;

s207: obtaining the angle of the pointer line segment according to the slope angle of the line segment formed by the start scale position and the circle center and the slope angle of the line segment formed by the end scale position and the circle center;

s208: and calculating the scale pointed by the pointer through a preset calculation formula to obtain the reading.

Calibrating the effective dial area: marking a starting scale point, an ending scale point and a correction point position on the circular dial (the circular dial), and inputting a starting position value num_bEnd position scale value num_eAnd the scale value (circular dial) adjust of the correction point, and record all the calibrated and input data. The position of the center of the circle (circular dial) is corrected and recorded. The midperpendicular of the start and end tick points is calculated. And calculating the position of the circle center on the middle vertical line and the effective dial area through the position and the scale value of the correction point.

Reading the dial plate reading: all line segments are detected within the active dial area. And setting rule parameters, carrying out regularized selection on the detected line segments, and filtering out the meter pointer line segments.

Circular dial plate pointer filters rule specifically includes:

the slope angle of the pointer segment is kept between the slope angles of the two end points and the circle center, kp belongs to [ kbegin-epsc, kend + epsc ], and epsc represents precision

The ratios for which the length of the pointer segment must be greater than the radius r represent the magnification, i.e.:

r_p＞r*ratios (3-1)

for a circular dial, the instrument is processed using the following algorithm:

and merging the regularized pointer line segments.

For any two straight lines, when the slope angle kp 1-kp 2 ∈ [ -eps, + eps ], then the two straight lines merge. eps stands for precision

And reading the reading corresponding to the pointer by an angle discrimination method.

Slope angle theta of line segment formed by starting scale position and circle center_bThe slope angle theta of the line segment formed by the ending scale position and the circle center_eAngle theta of pointer line segment_pBy the calculation formula:

<math> <mrow> <msub> <mi>P</mi> <mi>value</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mo>(</mo> <msub> <mi>&theta;</mi> <mi>p</mi> </msub> <mo>-</mo> <msub> <mi>&theta;</mi> <mi>b</mi> </msub> <mo>)</mo> </mrow> <mrow> <mo>(</mo> <msub> <mi>&theta;</mi> <mi>e</mi> </msub> <mo>-</mo> <msub> <mi>&theta;</mi> <mi>b</mi> </msub> <mo>)</mo> </mrow> </mfrac> <mo>*</mo> <mrow> <mo>(</mo> <msub> <mi>num</mi> <mi>e</mi> </msub> <mo>-</mo> <msub> <mi>num</mi> <mi>b</mi> </msub> <mo>)</mo> </mrow> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>3</mn> <mo>-</mo> <mn>2</mn> <mo>)</mo> </mrow> </mrow> </math>

and calculating the scale pointed by the pointer to obtain the reading.

FIG. 3 is a flowchart of an embodiment of square dial identification of the method for automatically reading a pointer instrument using video according to the present invention.

S301: acquiring a starting scale point and an ending scale point from an effective area of an input image calibration dial, and recording a scale value of a starting position and a scale value of an ending position;

s302: detecting all line segments in the effective area of the dial plate;

s303: setting rule parameters, carrying out regularized selection on the detected line segments, and filtering out an instrument pointer;

s304: projecting the regular pointer line segment to find a dial scale area;

s305: downwards translating a pointer area of the lookup instrument through the dial area boundary parameter;

s306: carrying out binarization processing on the pointer area, and finding the position of the pointer by a projection method;

s307: and calculating the scale pointed by the pointer through a distance discrimination method according to the starting position scale value, the ending position scale value and the pointer position to obtain the reading.

Calibrating the effective dial area: marking the start scale point and the end scale point on the square dial plate, and inputting the start position value num_bEnd position scale value num_e。

Reading the dial plate reading: all line segments are detected within the active dial area. And setting rule parameters, carrying out regularized selection on the detected line segments, and filtering out the meter pointer line segments.

The square dial pointer filtering rules are as follows: the pointer segment slope angle is kept at kp ∈ [90-epsr, 90+ epsr ] in the vertical direction, which represents precision.

For a square dial, the following algorithm is used to process the meter:

and performing projection processing on the regular pointer line segment to find a dial scale area.

And projecting the pointer line segment to the y-axis direction, searching the maximum value on the projection array, and searching the boundary of the dial plate area by up-down diffusion.

And searching for the pointer area through the dial area. And (4) translating downwards to find the pointer area of the instrument through the dial area boundary parameters.

And (3) carrying out binarization processing on the pointer region, finding the position of the pointer by a projection method, carrying out binarization on the pointer region by using an otsu algorithm, then projecting the pointer region onto an x axis, and searching the position of the pointer in the x direction by an extreme method.

And reading the reading corresponding to the pointer by a distance discrimination method. Starting position xbegin, scale value num_bEnd position xend, scale value num_ePointer position xcur, by the formula:

$P_{\mathrm{value}}=\frac{(x_{\mathrm{cur}}-x_{\mathrm{begin}})}{(x_{\mathrm{end}}-x_{\mathrm{begin}})}*({\mathrm{num}}_e-{\mathrm{num}}_b)---(3-3)$

and calculating the scale pointed by the pointer to obtain the reading.

Fig. 4 is a schematic diagram of the device for automatically reading the pointer instrument by using video according to the present invention, which comprises:

the image marking unit is used for marking the effective area of the dial plate from the input image, acquiring a starting scale point and an ending scale point, and recording a scale value of the starting position and a scale value of the ending position;

the line segment detection unit is connected with the image marking unit and is used for detecting all line segments in the effective area of the dial plate;

the pointer filtering unit is connected with the line segment detecting unit and used for setting rule parameters, performing regularized selection on the detected line segments and filtering out the meter pointers;

and the operation reading unit is connected with the pointer filtering unit and used for reading the scale pointed by the instrument pointer by adopting a corresponding algorithm according to the type of a dial plate to obtain a reading.

Fig. 4 corresponds to fig. 1, the various units of which operate in the same way as in the method.

Fig. 5 is a schematic diagram of an embodiment of the device for automatically reading a pointer instrument by using video.

As shown in fig. 5, includes: the circle center determining unit is connected between the image marking unit and the line segment detecting unit and used for calculating the perpendicular bisector of the starting scale point and the ending scale point; and calculating the position of the center of the circle on the middle vertical line by correcting the position of the point and the scale value.

As shown in fig. 5, the arithmetic reading unit includes:

the pointer merging unit is used for merging two straight lines when the slope angle kp 1-kp 2 belongs to [ -eps, + eps ] of the pointer line segment after the regularization;

the angle calculation unit is used for acquiring the angle of the pointer line segment according to the slope angle of the line segment formed by the starting scale position and the circle center and the slope angle of the line segment formed by the ending scale position and the circle center;

and the disc pointing unit is used for calculating the scale pointed by the pointer through a preset calculation formula.

As shown in fig. 5, the arithmetic reading unit includes:

the scale projection unit is used for performing projection processing on the regularized pointer line segment and searching a dial scale area;

the area searching unit is used for downwards translating and searching a pointer area of the instrument through the boundary parameters of the dial area;

the pointer searching unit is used for carrying out binarization processing on the pointer area and finding the position of the pointer by a projection method;

and the square disc pointing unit is used for calculating the scale pointed by the pointer through a distance discrimination method according to the starting position scale value, the ending position scale value and the pointer position.

As shown in fig. 5, the line segment detecting unit includes:

the Gaussian processing unit is used for carrying out Gaussian smoothing on the effective area to obtain a smooth image;

a gradient calculation unit for calculating a gradient and an edge line angle of the smoothed image;

the search list unit is used for establishing a search list and putting the position coordinates of each point into the corresponding BIN according to the gradient amplitude of each point in the order of big to small;

an auxiliary identification unit for creating an auxiliary image for temporarily recording identification information, setting the gradient amplitude smaller than theta as used, and setting the other as unused, wherein,

q is the pixel quantization precision, τ is the edge similarity angle precision;

and a pixel searching unit for starting searching according to the pixels in the searching list, wherein the searching rule is that the searching rule starts from the maximum gradient value in the BIN and the mark is unused.

The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims (8)

1. A method for automatically reading a pointer instrument using video, comprising:

acquiring a starting scale point and an ending scale point from an effective area of an input image calibration dial, and recording a scale value of a starting position and a scale value of an ending position;

detecting all line segments in the effective area of the dial plate;

setting rule parameters, carrying out regularized selection on the detected line segments, and filtering out an instrument pointer;

reading the scale pointed by the instrument pointer by adopting a corresponding algorithm according to the type of the dial plate to obtain a reading;

when the dial is rectangular, reading the scale pointed by the instrument pointer by adopting a corresponding algorithm, wherein the step comprises the following steps of:

projecting the regular pointer line segment to find a dial scale area;

downwards translating a pointer area of the lookup instrument through the dial area boundary parameter;

carrying out binarization processing on the pointer area, and finding the position of the pointer by a projection method;

and calculating the scale pointed by the pointer through a distance discrimination method according to the starting position scale value, the ending position scale value and the pointer position.

2. The method for automatically reading a pointer instrument with video according to claim 1, wherein when the dial is of the type of circular dial, the step of calibrating the active area of the dial is followed by the step of detecting all line segments within the active area of the dial, and comprises:

calculating a perpendicular bisector of the start tick mark and the end tick mark;

and calculating the position of the center of the circle on the middle vertical line by correcting the position of the point and the scale value.

3. The method of claim 2, wherein the step of reading the scale pointed to by the meter pointer using a corresponding algorithm comprises:

for the pointer line segment after the regularization, when the slope angle kp 1-kp 2 belongs to [ -eps, + eps ], then the two straight lines are merged; wherein eps represents precision, kp1 and kp2 represent slope angles of two straight lines, respectively;

obtaining the angle of the pointer line segment according to the slope angle of the starting scale point and the circle center forming the line segment and the slope angle of the ending scale point and the circle center forming the line segment;

and calculating the scale pointed by the pointer through a preset calculation formula.

4. The method of automatically reading a pointer instrument with video as claimed in claim 3, wherein the step of detecting all line segments within the active area of the dial plate comprises:

performing Gaussian smoothing on the effective region to obtain a smooth image;

calculating the gradient and the edge line angle of the smooth image;

establishing a search list, and putting the position coordinates of each point into a corresponding BIN according to the gradient amplitude of each point in a descending order;

creating an auxiliary image for temporarily recording the identification information, setting the gradient amplitude smaller than theta as used, and setting the other as unused, wherein,q is the pixel quantization precision, τ is the edge similarity angle precision;

the search is started by the pixel in the search list, starting with the largest gradient value in BIN and marked as unused.

5. An apparatus for automatically reading a pointer instrument using video, comprising:

the image marking unit is used for marking the effective area of the dial plate from the input image, acquiring a starting scale point and an ending scale point, and recording a scale value of the starting position and a scale value of the ending position;

the line segment detection unit is connected with the image marking unit and is used for detecting all line segments in the effective area of the dial plate;

the pointer filtering unit is connected with the line segment detecting unit and used for setting rule parameters, performing regularized selection on the detected line segments and filtering out the meter pointers;

the operation reading unit is connected with the pointer filtering unit and used for reading the scale pointed by the instrument pointer by adopting a corresponding algorithm according to the type of a dial plate to obtain a reading;

wherein, the calculation reading unit includes:

the scale projection unit is used for performing projection processing on the regularized pointer line segment and searching a dial scale area;

the area searching unit is used for downwards translating and searching a pointer area of the instrument through the boundary parameters of the dial area;

the pointer searching unit is used for carrying out binarization processing on the pointer area and finding the position of the pointer by a projection method;

and the square disc pointing unit is used for calculating the scale pointed by the pointer through a distance discrimination method according to the starting position scale value, the ending position scale value and the pointer position.

6. The device for automatically reading a pointer instrument by using video as claimed in claim 5, wherein when the dial is a circular dial, the device comprises:

the circle center determining unit is connected between the image marking unit and the line segment detecting unit and used for calculating the perpendicular bisector of the starting scale point and the ending scale point; and calculating the position of the center of the circle on the middle vertical line by correcting the position of the point and the scale value.

7. The device for automatically reading the pointer instrument by using the video as claimed in claim 6, wherein the arithmetic reading unit comprises:

the pointer merging unit is used for merging two straight lines when the slope angle kp 1-kp 2 belongs to [ -eps, + eps ] of the pointer line segment after the regularization; wherein eps represents precision, kp1 and kp2 represent slope angles of two straight lines, respectively;

the angle calculation unit is used for acquiring the angle of the pointer line segment according to the slope angle of the start scale point and the slope angle of the line segment formed by the circle center and the slope angle of the end scale point and the circle center;

and the disc pointing unit is used for calculating the scale pointed by the pointer through a preset calculation formula.

8. The apparatus for automatically reading a pointer instrument using video as claimed in claim 7, wherein the line segment detecting unit comprises:

the Gaussian processing unit is used for carrying out Gaussian smoothing on the effective area to obtain a smooth image;

a gradient calculation unit for calculating a gradient and an edge line angle of the smoothed image;

the search list unit is used for establishing a search list and putting the position coordinates of each point into the corresponding BIN according to the gradient amplitude of each point in the order of big to small;

an auxiliary identification unit for creating an auxiliary image for temporarily recording identification information, setting the gradient amplitude smaller than theta as used, and setting the other as unused, wherein,q is the pixel quantization precision, τ is the edge similarity angle precision;

and a pixel searching unit for starting searching according to the pixels in the searching list, wherein the searching rule is that the searching rule starts from the maximum gradient value in the BIN and the mark is unused.