Oil tank measurement method and system based on laser point cloud analysis

Abstract

The invention provides an oil tank measurement method and system based on laser point cloud analysis, comprising: acquiring point cloud data inside an oil tank, which is collected by a laser measurement device; separating point cloud data of a main body of the oil tank from point cloud data of a plug, to acquire the point cloud data of the main body of the oil tank; calculating, based on Gauss mapping, an axis for the point cloud data of the main body of the oil tank; determining any one first plane perpendicular to the axis, and projecting a point cloud of the main body onto the first plane to obtain a point cloud of a projected cross-section of the tank body on the first plane; multi-segment fitting the point cloud of the projected cross-section of the tank body; and calculating a volume according to a result of multi-segment fitting.

Claims

1. An oil tank measurement method based on laser point cloud analysis, comprising steps of: positioning a laser measurement device at an opening of an oil tank; scanning the oil tank to acquire point cloud data inside the oil tank using the laser measurement device and displaying a scanned image of an inside of the oil tank; separating point cloud data of a main body of the oil tank from point cloud data of a plug, to acquire the point cloud data of the main body of the oil tank; calculating, based on Gauss mapping, an axis for the point cloud data of the main body of the oil tank by: a) calculating a unit normal vector for each point in the point cloud data of the main body, translating a starting point of a normal vector to the origin of coordinates and forming a unit sphere by an ending point of the normal vector; b) plane fitting ending points of all normal vectors to obtain a plane Φ1; c)calculating a distance di (i=1,2, . . . n) from the ending point of each normal vector to the plane Φ1; d) setting a threshold ω1, deleting the normal vectors over a threshold ω1 distance if di>ω1, and fitting ending points of the remaining normal vectors again to obtain a plane Φ2; e) calculating a distance di′ from the ending points of all the remaining normal vectors to the plane Φ2 , and returning to the step d) until the distance from the ending points of all the remaining normal vectors to the plane is less than ω1 wherein a fit plane Φj is obtained which has a unit normal vector nj; and f) selecting one plane perpendicular to the unit normal vector nj, projecting the point cloud data of the main body onto the plane, and fitting the obtained circle center, wherein the axis passes through the circle center and is in a same direction as the unit normal vector nj; determining any one first plane perpendicular to the axis, and projecting the point cloud data of the main body onto the first plane to obtain a point cloud of a projected cross-section of the tank body on the first plane; multi-segment fitting the point cloud of the projected cross-section of the tank body; and calculating a volume according to a result of multi-segment fitting.

2. The oil tank measurement method based on the laser point cloud analysis according to claim 1, wherein multi-segment fitting the point cloud of the projected cross-section of the tank body is specifically as follows: uniformly dividing, by centering the axis and at a certain angle, the point cloud of the projected cross-section of the tank body into 4N segments, N=1,2,3 . . . ; and fitting each segment of the point cloud of the projected cross-section of the tank body, respectively.

3. The oil tank measurement method based on the laser point cloud analysis according to claim 2, wherein, for the fitting of the point cloud of the projected cross-section of the tank body, quadratic curve fitting is used.

4. The oil tank measurement method based on the laser point cloud analysis according to claim 1, wherein calculating a volume according to a result of multi-segment fitting specifically comprises: acquiring auxiliary calculation data measured in advance, wherein the auxiliary calculation data comprises the length of the main body of the oil tank and the height of the plug; and calculating the volume according to the result of multi-segment fitting and the auxiliary calculation data.

5. The oil tank measurement method based on the laser point cloud analysis according to claim 1, further comprising a step of: acquiring a unit distance in a length direction, and containing the volume per unit distance in the volume calculation result.

6. The oil tank measurement method based on the laser point cloud analysis according to claim 5, further comprising a step of: outputting a measurement result, wherein the measurement result comprises submitter, name of measuring instrument, model/specification, tank No., manufacturer, verification/calibration basis, conclusion, date, volume per unit distance, horizontal diameter, vertical diameter, length of the tank body, external height, approximate cylindrical volume, plug volume, and total volume.

7. The oil tank measurement method based on the laser point cloud analysis according to claim 6, wherein the measurement result is output in Excel format.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a flowchart of an oil tank measurement method based on laser point cloud analysis in an exemplary embodiment;

(2) FIG. 2 is a segmented view of the point cloud of the projected cross-section of the tank body in an exemplary embodiment; and

(3) FIG. 3 is a schematic block diagram of an oil tank measurement system based on laser point cloud analysis in an exemplary embodiment.

DETAILED DESCRIPTION OF THE PRESENT INVENTION

(4) The technical solutions of the present invention will be further described with reference to the accompanying drawings by specific embodiments of the present invention. However, the present invention is not limited to those embodiments.

(5) At present, the capacity measurement of oil tanks has disadvantages of low measurement speed, high labor intensity, low degree of automation, waste of water, and lack of energy saving awareness. To solve those problems, the present invention provides an oil tank measurement method based on laser point cloud analysis.

(6) The oil tank measurement method based on laser point cloud analysis in an embodiment of the present invention will be further explained and described with reference to the accompanying drawings.

(7) FIG. 1 is a flowchart of an oil tank measurement method based on laser point cloud analysis in an exemplary embodiment. As shown in FIG. 1, the oil tank measurement method based on laser point cloud analysis comprises steps S101 to S106.

(8) S101: Point cloud data inside an oil tank, which is collected by a laser measurement device, is acquired.

(9) The collected data information includes size information inside the oil tank, washboard and guide tubes.

(10) The process of data collection by the laser measurement device is as follows: the laser measurement device is hung upside down at the opening of the oil tank and well fixed, and then the leveling bubble is adjusted to the middle position. Then, measurement parameter data for inclination and rotation, for example, measurement speed, measurement interval, the number of measurement steps, upper and lower limits, starting point of measurement, are set in the control system. Then, the measurement is started. At the end of measurement, the data is uploaded to a computer.

(11) In the computer, the collected data may be overall viewed by man-machine interaction to obtain the distribution of noise point clouds and stray points. The noise point clouds may be selected, abandoned, deleted and hidden by a rectangular box. And, the washboard inside the tank body and the stray points may be eliminated.

(12) S101 may further comprise checking the integrity of the point cloud data, in which the point cloud data is rotated up and down, left and right, and horizontally, to check the integrity of the collected point cloud data information of the oil tank.

(13) It is to be noted that, by the laser measurement device, the oil tank may be scanned at 360°, and the scanned image of the inside of the tank body is clear, which is convenient for the user to view.

(14) S102: Point cloud data of a main body of the oil tank is separated from point cloud data of a plug, to acquire the point cloud data of the main body of the oil tank.

(15) S103: Based on Gauss mapping, an axis is calculated for the point cloud data of the main body of the oil tank.

(16) Specifically, the step S103 comprises:

(17) S21: calculating a unit normal vector for each point in the point cloud of the main body, translating a starting point of a normal vector to the origin of coordinates and forming a unit sphere by an ending point of the normal vector;

(18) S22: plane fitting ending points of all normal vectors to obtain a plane Φ1;

(19) S23: calculating a distance di (i=1,2, . . . n) from the ending point of each normal vector to the plane Φ1;

(20) S24: setting a threshold ω1, deleting normal vectors over a threshold ω1 distance if di>ω1, and fitting ending points of the remaining normal vectors again to obtain a plane Φ2;

(21) S25: calculating a distance di′ from the ending points of all the remaining normal vectors to the plane Φ2, and returning to the step S24 until the distance from the ending points of all the remaining normal vectors to the plane is less than ω1, wherein a fit plane Φj is obtained which has a unit normal vector nj; and

(22) S26: selecting one plane perpendicular to the unit normal vector nj, projecting the point cloud of the main body onto the plane, and fitting the obtained circle center, wherein the axis passes through the circle center and is in a same direction as the unit normal vector nj.

(23) First, a unit normal vector is calculated for each point in the point cloud of the main body, a starting point of a normal vector is translated to the origin of coordinates and a unit sphere is formed by an ending point of the normal vector; ending points of all normal vectors are plane fit to obtain a plane Φ1; normal vectors over a threshold ω1 distance are deleted; ending points of the remaining normal vectors are fit again to obtain a plane Φ2; normal vectors over the threshold ω1 distance are deleted again based on the newly generated plane Φ2, and the new plane is then fit; the above process is repeated until the distance from the ending points of all the remaining normal vectors to the plane is less than ω1, wherein a fit plane Φj is obtained which has a unit normal vector nj; one plane perpendicular to the unit normal vector nj is selected, the point cloud of the main body is projected onto the plane, and the obtained circle center is fit, wherein the axis passes through the circle center and is in a same direction as the unit normal vector nj.

(24) S104: Any one first plane perpendicular to the axis is determined, and a point cloud of the main body is projected onto the first plane to obtain a point cloud of a projected cross-section of the tank body on the first plane.

(25) S105: The point cloud of the projected cross-section of the tank body is multi-segment fit.

(26) Specifically, multi-segment fitting the point cloud of the projected cross-section of the tank body is specifically as follows:

(27) uniformly dividing, by centering the axis and at a certain angle, the point cloud of the projected cross-section of the tank body into 4N segments, N=1,2,3 . . . ; and

(28) fitting each segment of the point cloud of the projected cross-section of the tank body, respectively.

(29) Further, for the fitting of the point cloud of the projected cross-section of the tank body, quadratic curve fitting is used.

(30) Specifically, since it is difficult to perform overall fitting because the projected cross-section of the tank body is irregular, segmented fitting is used. The point cloud of the cross-section is uniformly divided, at a certain angle, into 4N segments, N=1,2,3 . . . . Then, each segment of the point cloud is fit, respectively. There are many ways of fitting, for example, quadratic curve fitting, spline curve fitting, etc. In addition, in consideration of the accuracy of fitting, quadratic curve fitting is used. If the amount of data is sufficient, a greater number of segments to be fit results in a fit curve that is closer to the reality and thus a higher accuracy of fitting. FIG. 2 is a segmented view of the point cloud of the projected cross-section of the tank body in an exemplary embodiment.

(31) S106: A volume is calculated according to a result of multi-segment fitting.

(32) Specifically, first, the cross-sectional area is calculated according to the result of multi-segment fitting, and the capacity of the oil tank is calculated in combination with the data in the length direction.

(33) Further, calculating a volume according to a result of multi-segment fitting specifically comprises:

(34) acquiring auxiliary calculation data measured in advance, wherein the auxiliary calculation data comprises the length of the main body of the oil tank and the height of the plug; and calculating the volume according to the result of multi-segment fitting and the auxiliary calculation data.

(35) Specifically, the auxiliary calculation data measured in advance is entered in the system, for example, the length of the main body of the oil tank, the height of the plug and the external height. Such data provides basic information support for the calculation of the horizontal and vertical diameters of the tank body and the volume of each part.

(36) Further, the oil tank measurement method based on laser point cloud analysis further comprises a step of: acquiring a unit distance in a length direction, and containing the volume per unit distance in the volume calculation result.

(37) Specifically, the calculation spacing is defined. It may be input by an operating interface. Usually, 1 mm or 10 mm may be input, which denote a millimeter volume table or a centimeter volume table in the output template, respectively.

(38) Further, the oil tank measurement method based on laser point cloud analysis further comprises a step of:

(39) outputting a measurement result, wherein the measurement result comprises submitter, name of measuring instrument, model/specification, tank No., manufacturer, verification/calibration basis, conclusion, date, volume per unit distance, horizontal diameter, vertical diameter, length of the tank body, external height, approximate cylindrical volume, plug volume, and total volume.

(40) It is to be noted that the content contained in the measurement result is not limited thereto and may be adjusted according to actual situations.

(41) Further, the measurement result is output in Excel format.

(42) Compared with the prior art, the oil tank measurement method in this embodiment causes no water consumption; by the automatic measurement, the verification efficiency is improved significantly when compared with the conventional methods, and the shutdown time of enterprises is reduced; there is no need for the operators to implement measurement inside the tank, no standard metal measuring instruments are required, and the measurement may be realized outside the oil tank, so the labor intensity is decreased greatly; and the used laser measurement device is of an intrinsically safe explosion-proof design, so it may be used in oil and gas environments, and the safety of operators is ensured.

(43) FIG. 3 is a schematic block diagram of an oil tank measurement system based on laser point cloud analysis in an exemplary embodiment. The oil tank measurement system based on laser point cloud analysis comprises:

(44) an acquisition module 31, configured to acquire point cloud data inside an oil tank, which is collected by a laser measurement device;

(45) a separation module 32, configured to separate point cloud data of a main body of the oil tank from point cloud data of a plug, to acquire the point cloud data of the main body of the oil tank;

(46) an axis calculation module 33, configured to calculate, based on Gauss mapping, an axis for the point cloud data of the main body of the oil tank;

(47) a projection module 34, configured to determine any one first plane perpendicular to the axis, and project a point cloud of the main body onto the first plane to obtain a point cloud of a projected cross-section of the tank body on the first plane;

(48) a fitting module 35, configured to multi-segment fit the point cloud of the projected cross-section of the tank body; and

(49) a volume calculation module 36, configured to calculate a volume according to a result of multi-segment fitting.

(50) The oil tank measurement system based on laser point cloud analysis in this embodiment corresponds to the oil tank measurement method based on laser point cloud analysis shown in FIG. 1. The detailed description of the system may refer to the description of the embodiment of the method and will not be repeated here.

(51) The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or supplements or replacements in a similar way of the described specific embodiments without departing from the spirit of the present invention or beyond the scope defined by the appended claims.