191 results about "Circle fitting" patented technology

The invention relates to a method for measuring the diameter and the straightness accuracy parameters of a seamless round steel pipe. The method includes the steps as follows: accomplishing the calibration of parameters of a multi-linear structured light vision sensor formed by one video camera and multi-line laser projectors; distributing the multi-linear structured light vision sensors on the periphery of the seamless round steel pipe to be tested; respectively controlling the laser projectors to project structured light plane onto the measured section of the seamless round steel pipe by a computer, collecting the images of laser stripes on the surface of the seamless round steel pipe and processing the images of the laser strip by a video camera, and calculating the three-dimensional coordinates of different sections according to a measurement model; determining the elliptical center of each section of the space, realizing the measurement of the straightness accuracy of the seamless round steel pipe and establishing the dynamic and virtual central axis of the seamless round steel pipe; establishing the reference plane of the dynamic and virtual projection of ellipse of the section of the space; and front projecting the ellipse of the section of the space onto the reference plane of the dynamic and virtual projection, and carrying out the operation of circle fitting on the reference plane, thereby obtaining the section circle of the steel pipe. The method can realize the on-line, real-time, automatic and noncontact measurement of the diameter and the straightness accuracy parameters of the seamless round steel pipe.

The invention discloses a spatial circle geometrical parameter dual-eye solid visual measurement method suitable for non-contact type online measurement of the spatial circle geometrical parameter of an industrial product, aiming at overcoming the existing problems of low measurement precision, slow measurement speed and low automation degree. The method is divided into a calibrating stage, an image processing stage and a spatial circle fitting stage. A neural network technique is firstly adopted to carry out the calibration of a camera; a dual-camera is utilized to carry out the extraction of sub-pixel image edges of the spatial circle, a simple algorithm is developed based on the image gray gradient distribution characteristic so as to realize the match of edge points, the practical 3D spatial coordinates of the circle edges are then obtained, and the geometrical parameters of the spatial circle comprise radius, position of center of circle and orientation of the plane where the spatial circle is arranged are gained by the fitting of the spatial circle with corresponding mathematical geometrical knowledge. The spatial circle geometrical parameter dual-eye solid visual measurement method has fast image processing speed and high automation degree, and has relative measurement error of the spatial circle of superior to plus or minus 0.6% when the plane where the spatial circle is arranged has an angle of about 50 DEG with the image plane.

The invention adopts the between-class square error and the gray scale information to realize the rapid positioning of the inside and outside boundaries of the iris. Firstly, the interested pupil region is extracted through blocking; the between-class square error is adopted to obtain the pupil binary threshold for the extracted pupil region; then the inside boundary of the iris is positioned accurately through the searching of the boundary points and the curve fitting; the interested region of the iris outside boundary is extracted according to the prior knowledge and the pupil position parameter; the selected region is processed with median filtering and first-order gradient conversion; the iris outside boundary is determined through the local gradient integration method; finally, the iris outside boundary parameter is determined through the circle fitting. The method avoids the image binary from depending on the histogram; the image positioning time is greatly improved because only the interested region is processed; the whole image is not processed; the experiment indicates that the robustness and the positioning efficiency of the algorithm can satisfy the real-time processing requirement of the image.

The invention belongs to the field of computer numerical control, especially relating to a method fro controlling silicon wafer prealignment, comprising : a silicon wafer locating method comprising circular sampling of silicon wafer, data preprocessing, circle fitting and silicon wafer centering; and a gap locating method comprising gap rough locating, gap fine sampling, gap fitting and angling the gap center. The invention can largely improve the aligning accuracy.

The invention relates to a iris location method based on border point search and SUSAN edge detection, wherein the method using projection method first, to do and horizontal gray projection for binary iris image filled by spot, obtain rough center of pupil; through vertical scanning gray curve, to find a point which gray value is greater than a threshold value as internal border point of iris; circle fitting a series of the internal border points, thus obtaining location information of inner edge of the iris; Through SUSAN edge detection method, to cause the situation of obvious external border of the iris characteristic solved well, through direction detection, obtaining radius of external border of the iris, and location information of external border of the iris accurately; locating external border of the iris through the SUSAN edge detection method outside edge of the iris position is a feature of the invention, comparing with general two-step iris location method, positioning accuracy rate of the invention is enhanced by 5 percentage, and the speed of the same is enhanced by 40 percent.

The invention discloses a mechanical arm tool coordinate system self-calibration method, and belongs to the field of correlation technique of robot calibration. The method is applicable to an XYZR cartesian coordinate robot, and comprising the steps of utilizing a self point position of the robot for calibrating a tool center point; and through adjusting multiple points and corresponding multipleposition data of a tool fixedly arranged on a rotary shaft at the tail end of a mechanical arm and a top until the required points are recorded, circle fitting the recorded point position data to obtain a center coordinate, and obtaining a tool coordinate according to a coordinate geometry relationship. According to the mechanical arm tool coordinate system self-calibration method provided by theinvention, no external measurement tool is required, the tool coordinate system can be identified relative to a pose of a terminal coordinate system by utilizing self conditions, so that the method issimple and convenient for field application of the equipment party.

The invention relates to a binocular vision-based unmanned aerial vehicle aerial autonomous refueling fast docking navigation method and belongs to the machine vision and image processing field. The method includes the following steps that: a binocular camera system with an optical filter shoots a refueling taper sleeve with an optical marker lamp, so that a left image and a right image are obtained; gray processing, binarization and median-value filtering are performed on the left image and the right image sequentially, all communicated regions in the left image and the right image are searched and marked through adopting a regional growth method, and the centroids of each of the communicated regions are calculated and are adopted as feature points; improved haar wavelet transform is adopted to describe all the feature points, and the description vectors of all the feature points are obtained, the feature points of the left image and the right image are matched through adopting the minimum Euclidean distance method, so that the feature point pars of the left image and the right image are obtained; and the three-dimensional coordinates of the feature points are calculated by using a binocular vision principle, and the least squares method is adopted to perform spatial circle fitting on the three-dimensional coordinates, and the end surface circle curve of the refueling taper sleeve is obtained, and the center, normal vector and radius of the circle are obtained through calculation.

The invention relates to a docking ring circle center extraction method based on a close-range short-arc image. The docking ring circle center extraction method comprises the following steps: carrying out epipolar rectification on a binocular image containing a short-arc docking ring, dividing a threshold by using a standard deviation method, extracting a ring surface by means of regional attributes, and then extracting inner and outer edges; carrying out dense reconstruction on the extracted ring surface region by using a semi-global matching method to obtain three-dimensional point cloud in a camera coordinate system; calculating a ring surface normal vector of the point cloud, and projecting a coordinate in a plane vertical to the normal vector; distinguishing the inner and outer edges in a new plane, and respectively carrying out perfect circle fitting on the inner and outer edges; and finally, obtaining an accurate circle center coordinate of the docking ring by use of concentric ring constraint according to obtained centers and radiuses of the inner and outer edges. According to the docking ring circle center extraction method provided by the invention, from the characteristics of the docking ring, three-dimensional point cloud data are directly projected into a normal incidence spatial plan, ellipse fitting is simplified to circle fitting, and meanwhile, error points are eliminated by constraint conditions of inner and outer rings, and the docking ring circle center extraction method has the advantages of accurate result and small calculated amount.

The invention relates to a notch circular arc part dimension visual detection method based on self-adapting region division. The method comprises the following steps of (1) image binaryzation; (2) notch circular arc contour outline extraction; (3) circular arc point region self-adapting division; (4) inner and outer circular arc point separation; and (5) inner and outer circle fitting, calculation of the distance between O<in> and O<out> (namely the concentricity), and result output. A notch circular arc part pattern is obtained in the following way that the contour outline of a detected notch circular arc part is illuminated by a light source lamp house; an industrial camera shoots the illuminated notch circular arc part and enables the part to be completely shown in an image collected by the industrial camera; and image data is transmitted to a computer through a data wire. The notch circular arc part dimension visual detection method based on the self-adapting region division provided by the invention has the advantages that the extraction is performed by aiming at parts with a notch smaller than the semicircular arc dimension, incomplete inner and outer circular arc part or inner and outer circle arc points; self-adapting regulation can be realized on the concrete image condition; relevant parameters are output; and the real-time continuous and accurate industrial production detection requirements can be met.

The invention discloses a static contact angle calculation method in the technical field of material performance test. The method includes: simulating to generate droplet margins of different sizes and contact angles on the basis of the Young-Laplace equation; calculating the contact angles of the droplet margins generated by simulation to obtain relation among the calculated contact angles, the droplet sizes and the real contact angles; taking a picture of a real droplet image and calculating the contact angle of the image; and using the interpolation or similar methods to obtain accurate contact angles according to the contact angle of the actual image and the relation among the calculated contact angles, the droplet sizes and the real contact angles. Errors brought by large droplet sizes and contact angles to the circle fitting method and theta / 2 method can be effectively reduced, calculation efficiency is improved, and programming difficulty is lowered.

The invention relates to a measurement method and system of a non-contact tube-class object. The measurement method comprises the following steps of: step 1, tube-class object image acquisition; step 2, tube-class object three-dimensional model reconstruction; and step 3, tube-class object center line extraction: encapsulation, triangular patch processing and materialization are performed on the three-dimensional point cloud data of a tube-class object obtained in the second step, so that a three-dimensional solid model of the tube-class object can be obtained, and center line data are extracted from the three-dimensional solid model, or piecewise cutting is performed on the three-dimensional point cloud data obtained in the second step, circle fitting is performed on points on the cross sections of the three-dimensional point cloud data, so that the center of each cross section circle is obtained, and the centers of the cross section circles are a series of discrete points on the center line of the tube-class object, curve-fitting is performed through using the least square method and based on the discrete points, so that the center line data can be extracted.

The invention discloses a measuring method for the relative liquid level position of silicon melt in a Czochralski silicon single crystal furnace. Laser reflected by the liquid level of the silicon melt is captured by a CCD (charge coupled device) camera, so that a laser image can be obtained; the circle center of the laser image is obtained through the perfect circle fitting of edge pixel coordinates; and the y coordinate difference between the circle center of the perfect circle fitting diagram of a laser image of any liquid level position and the circle center of the fitting diagram of the initial liquid level position indicates the relative height of the liquid level position. The measuring method provided by the invention is a simple and effective method for precise real-time measurement and control of the relative changes of the liquid level position of the silicon melt in a quartz crucible during the silicon single crystal growth.

The invention relates to a method and a device for automatically determining the scanning range of the petrous bone. The method comprises the following steps: determining the edge of the skull according to a head part positioning image; implementing head part circle fitting according to the determined edge of the skull to obtain a skull ring; determining an initial interesting area of the petrous bone according to the skull ring; using the initial interesting area as a scanning range of the petrous bone. According to the method and the device disclosed by the embodiment of the invention, the scanning range of the petrous bone can be automatically arranged; the time-costing labor-consuming manual operation and the resultant defects of artificial errors, low efficiency and the like are avoided; the workflow is simplified; the efficiency is improved; in addition, the eye positions can be automatically detected, and the petrous bone scanning range excluding the eyes can be automatically set through the method and the device.

The invention provides a roller tractor turning control method and device. The method comprises the steps that the point coordinates of multiple points on a track circle formed by moving of a roller tractor for at least one circle at a constant speed are acquired, and the track circle is formed in a way that the steering wheel of the roller tractor is rotated to a preset position and then maintained unchanged in moving; fitting calculation of the point coordinates of multiple points is performed by utilizing a circle fitting method, and the moving track circle radius R<0> of the roller tractor is determined; moving turning radius R=R<0>+L<tw> / 2 of the roller tractor is determined according to front wheel distance L<tw> of the front wheels of both sides of the roller tractor and the track circle radius R<0>; and turning control of the roller tractor is performed according to the turning radius R=R<0>+L<tw> / 2. According to the roller tractor turning control method and device, the problems in the prior art that measurement technological means are poor and measurement precision is low can be effectively overcome, turning control precision of the roller tractor can be guaranteed, and enhancement of working efficiency of the roller tractor is facilitated.

The invention relates to a method for denoising three-dimensional laser scanning point cloud data of a shield tunnel. According to the method, the technical problems of complicated steps, low denoising precision and the like of an existing method for denoising scanning point data are mainly solved. The method provided by the invention is characterized by comprising the following steps: acquiring three-dimensional laser scanning data of the shield tunnel; S1, reading scanning point cloud data, marking an each cycle of scanning point data set as E0; S2, establishing a Cartesian rectangular coordinate system and partitioning scanning point quadrants; S3, extracting the scanning point cloud data; S4, carrying out preliminary elimination on noisy points according to structural features of the shield tunnel and scanning features of a three-dimensional laser scanner based on X values or Y values of different quadrants and the distance change trend; S5, carrying out circle fitting secondary elimination on the scanning point data subjected to preliminary elimination of the noise points; S6, carrying out final denoising of the three-dimensional laser scanning point cloud data by an ellipse fitting method performed on the basis of a geometrical distance least square method; S7, repeating the steps of S2, S3, S4, S5 and S6, and treating all point cloud data of all scanning cycles; S8, outputting effective point cloud data of the scanning points.

The invention provides a circular center detection method based on Gaussian fitting sub-pixel edge detection and an improved RANSAC algorithm, belonging to the field of aircraft ground simulation. Themethod comprises the follow concrete steps: an original image is matched by a model and RIO processing is conducted on the image; edge pixel extraction and thinning is carried out; sub-pixel edge location based on Gaussian fitting is carried out; an improved RANSAC circle fitting algorithm is provided; the coordinates of the center of a circle are calculated. The invention roughly locates the position of the circular marker through a model matching mode, and performs RIO processing to reduce the size of image processing. Gaussian fitting improves the accuracy and stability of the algorithm. The improved RACNAC algorithm reduces the sampling times, optimizes and screens the candidate circles, reduces the running time and improves the accuracy and stability.

The invention discloses an installation error identification method based on a rotating shaft integration error measurement value. The method comprises the following steps of step1, taking the intersection point of a workbench rotation C rotation shaft axis and a workbench surface as an original point, and establishing a measured coordinate system; step2, through a measuring instrument, acquiringthe comprehensive error measurement value of the theoretical point and the actual point of the different rotation angles of a point P around a C shaft; step3, extracting the error projection point ofthe comprehensive error measurement value in an XOY measurement coordinate plane, and carrying out plane circle fitting to obtain the displacement error of a C shaft axis; and step4, extracting the error projection points of the comprehensive error measurement value in an XOZ measurement coordinate plane and a YOZ measurement coordinate plane, and carrying out straight line fitting and acquiring the perpendicularity errors of the C shaft, an X shaft and a Y shaft. In the invention, the four installation errors of the rotation shaft of a five-shaft numerical control machine tool can be conveniently, rapidly and accurately acquired, a data reference is provided for the assembling and debugging and the installation error of the rotation shaft of the five-shaft numerical control machine tool,and the method is suitable for all structural types of the five-shaft numerical control machine tools.

The invention discloses a method for detecting a rotating part of radiation therapy equipment. The detection method comprises the following steps of 1, fixing a laser sensor at a rotating axis of a rotatable part of the radiation therapy equipment, and respectively arranging three laser positioning lamps on two sides and the top of the laser sensor; 2, calculating intersection coordinates of cross curves of the laser positioning lamps by a computer according to laser signals received by the laser sensor; 3, rotating the rotatable part according to a preset angle; 4, repeating steps 2 and 3 twice at least, and calculating coordinates of a circle center and a radius according to the obtained at least three groups of intersection coordinates through a least square circle fitting equation; 5, detecting whether a difference value between an included angle between radii and a preset angle and difference values between the coordinates of the circle center and horizontal coordinates of the rotating axis are within a standard error range. The method has the beneficial effects that the method completely replaces a manual visual inspection method, the detection is faster, and detection results are more accurate.

The invention discloses a ground laser radar scanning-based automatic tree breast-height diameter calculation method. According to the method, the calculation speed and robustness of the algorithm are improved through trunk cutting and filtration; the condition that the breast-height diameters of the trees cannot be calculated as point clouds do not exist at the 1.3m places or the noise is too loud through automatic selecting point clouds to carry out fitting; the problem of incorrect circle fitting caused by the trunk growth angle problem is overcome through cylinder fitting; and the calculation precision is improved through a sliding window fitting cylinder. The method has nothing to do with the density, is insensitive to partial deficiency of the trunks and overcomes the problem that the long-distance trees are difficult to extract as the density is too low or loss is caused due to sheltering; and the method is independent of scanning station number, has no requirements for the fact whether the trunk point clouds are circular or not, and can achieve good effect even single-station scanning is carried out.

The invention discloses a beer bottleneck positioning method based on the visual sense of a machine. The method includes the steps: (1) carrying out threshold processing on a bottleneck image, and obtaining the substantial position of the center of circle of a bottleneck through a center of gravity method; (2) conducting radial scanning with the center of circle being a center of gravity, and obtaining external periphery points of the bottleneck image; (3) removing interference points by means of the change characteristics of distances between the center of gravity and all the periphery points; (4) dividing the periphery points into groups, freely selecting Nc groups of the periphery points, conducting circle fitting by the least square method, obtaining proportions [Eta]j of the number of the periphery points on fit circles to the total number of the periphery points, and saving all fitting results and the corresponding proportions [Eta]j; and (5) selecting the center of one fit circle as the center of the bottleneck, the proportion of the number of the periphery points on the fit circle to the total number of the periphery points being the highest. The method can accurately and quickly position the center of a bottleneck if a beer bottleneck image is not complete and continuous interference points exist, and is suitable for automatic detection of the quality of beer bottlenecks on a high speed automatic assembly line.

A related tower inclination measure method based on circle fitting comprises the following steps: 1) fitting plane equations of all planes corresponding to a polygonal tower; 2) obtaining vertex coordinates at high positions of tower marks at all layers according to the plane equations of all planes; 3) determining the center coordinates of all layers of the tower by employing a circle fitting method, namely determining the plane circle center approximate value of planes at all layers of the tower by utilizing the vertex coordinates of all layers, and then finding the centers of all layers of the tower by utilizing the circle fitting method; and 4) finally obtaining a space straight line with smallest error by utilizing the center coordinates of all layers, and taking the space straight line as an axis of the tower, so as to obtain the inclination degree and direction of the tower. Compared with the prior art, the method has the advantages of precise measuring, usage convenience, wide application scope and the like.

The invention discloses a position detection device and method for a steel roll system. The device comprises a total station, a reference landmark and a marker post, wherein the marker post comprisesa marker post arm, a prism, a slide block and a vacuum suction disc, the total station is bridged on one sides of the reference landmark and a to-be-measured roll shaft, the prism is fixedly connectedto the top of the marker post, the marker post arm can move up and down along the slide block and is positioned through a positioning pin, and the slide block is fixedly connected to the vacuum suction disc. The position detection device has the advantages that multiple parameter indexes of a single roll and the roll system can be simultaneously detected by virtue of the whole set of a measurement device; and by utilizing an adaptive threshold adjustment algorithm, the accuracy and fitting speed of a traditional spatial circle fitting algorithm are increased, and the device can be used for guiding the correction of the roll system, so that the productivity, quality and benefit of a product are greatly increased.

The invention discloses a cylindrical surface radius measuring method based on a three-dimensional vision. The cylindrical surface radius measuring method comprises the steps that a three-dimensional vision measuring system obtains three-dimensional point cloud data of a cylindrical surface, and the three-dimensional vision measuring system is composed of a projector and a camera; parallel planes are adopted to segment the three-dimensional point cloud data into two-dimensional plane data; an ordering algorithm based on angles is adopted to expand the two-dimensional plane data into a one-dimensional point sequence; over-limit median filtering and secondary filtering are carried out on the one-dimensional point sequence, so that the one-dimensional point sequence after noise removal is obtained; ellipse fitting and circle fitting are carried out according to the obtained one-dimensional point sequence, so that the magnitudes of radiuses at different cross sections of the cylindrical surface are obtained. According to the invention, the magnitudes of radiuses at different cross sections of a cylindrical part are obtained, and position error measurement and analysis can be directly carried out according to the calculated magnitudes of radiuses at different cross sections, so that the method is accurate, comprehensive, high in noise resistance capability and measurement precision, and can be widely applied to the field of computer vision detection.

A wafer prealignment control method comprises the following steps: a first step of providing a negative pressure absorption rotating device for pumping air between a tray and a wafer to form a vacuum state and driving the wafer to rotate, and a second step of providing a sensor used for collecting edge data of the wafer and recording rotating speed information of the rotating device. The wafer prealignment control method further comprises wafer positioning and wafer notch positioning, wherein the wafer positioning comprises the following steps: a first step of data acquisition, a second step of data preprocessing, and a third step of circle fitting to find the circle center position of the wafer. The wafer notch positioning comprises the following steps: a first step of notch coarse positioning, a second step of notch fitting, a third step of an iteration termination, a fourth step of finding a coordinate of the center according to the notch edge center, and the position and the angle of the wafer are adjusted.

The invention discloses a spherical pin element positioning algorithm based on point matching, which relates to the field of spherical pin element positioning. The spherical pin element positioning algorithm is disclosed to solve the problem of lack of accurate measurement of a position and an angle of a spherical pin chip at present. The spherical pin element positioning algorithm comprises the steps of: drawing a white circular image according to a radius size of a spherical pin and performing distance transformation; conducting edge detection on an original chip image and performing correlation filtering to obtain center coordinates of a solder ball point; selecting center coordinates of a solder ball at the peripheral endpoint of a chip for point matching, so as to obtain a rotary angle of the chip and deviation position coordinates of a chip entity center relative to the center of the chip image; fitting an edge image of the chip with a coordinate set shown in description by adopting a circle fitting method, so as to obtain a precise rotary angle of the chip and a precise deviation position coordinates of the chip entity center relative to the center of the chip image, thereby realizing the positioning of a spherical pin element. The spherical pin element positioning algorithm based on point matching is applied to the testing process of a spherical pin chip of a chip mounter.

The invention discloses an intelligent empty bottle detection robot bottle bottom positioning method. The method comprises the following steps of (1) preprocessing the image of a bottle bottom; (2) acquiring the approximate position of the circle center of the bottle bottom through the gravity method; (3) conducting the radial scanning with the center of gravity as the circle center, according to the geometric characteristics of anti-skid patterns on the bottle bottom, acquiring the position information of anti-skid lines in the image of the bottle bottom, introducing a weighting operator and removing interfering points; (4) grouping the edge points of anti-skid lines, optionally selecting edge points of Nc groups, and conducting the circle fitting through the variable-weight least squares method. The method overcomes the defect that the positioning effect is not accurate due to the uneven light transmittance of the bottle bottom. Meanwhile, when a glass bottle bottom is blocked by foreign matters or continuous interfering points exist on the glass bottle bottom, the center of the bottle bottom can be accurately and rapidly positioned. Therefore, the method is suitable for the quality automatic detection of glass bottle bottoms on a high-speed automatic production line.

The invention provides a stumpage breast height diameter measuring method based on an optics similar triangle method. The stumpage breast height diameter measuring method based on the optics similar triangle method relates to forest measurement and includes the following steps: a camera and a laser are used to create a stumpage breast height diameter measuring system; the camera is used to take photos of a to-be-tested stumpage to obtain images of the stumpage; and after the shot images of the stumpage are preprocessed, a point, namely a laser spots falling on a phase plane, (X', Y') is obtained. According to an optics similar triangle principle, the point (X', Y' ) that the laser spot falls on the phase plane carries out coordinate transformation towards a three-dimensional space to obtain a point (X, Y, Z). The point (X, Y, Z) projects towards a horizontal direction to obtain a point (X projection, Y projection) after being processed. A least square circle fitting method is adopted to carry out circle fitting on the point(X projection, Y projection), and then a round radius parameter after fitting is a breast height diameter of the to-be-tested stumpage. The stumpage breast height diameter measuring method based on the optics similar triangle method has the advantages that algorithm can be realized easily and can measure the stumpage breast height diameter quickly in non-contact mode.

The embodiment of the invention discloses a method and a device for extracting the point cloud data of an electric transmission line tower. The method comprises the following steps: estimating the tower top position point of the electric transmission line tower; estimating the tower arm height of the electric transmission line tower according to the height of the tower top position point, and dividing the electric transmission line tower into a tower top part and a tower body part according to the estimated tower arm height; extracting the point cloud data of the tower top part according to a circle fitting manner; and extracting the point cloud data of the tower body part according to a fitting manner towards an oriented bounding box (OBB). According to the method and the device for extracting the point cloud data of the electric transmission line tower provided by the embodiment of the invention, the accuracy of point cloud data extraction of the electric transmission line tower is improved.

The invention relates to a measuring method of oil drop sizes. The measuring method comprises the steps: manufacturing a two-phase flow fluid sampler, wherein a middle sampling shooting section is shaped like flat cuboid space and is made of transparent materials; connecting both ends of the two-phase flow fluid sampler to a to-be-measured pipeline by flanges, and arranging a high-speed camera lens opposite to two-dimensional surfaces formed by long edges and wide edges of the sampler, so as to shoot fluid in the sampler when oil-water two-phase flow flows across the middle sampling shooting section of the sampler; extracting oil drop edges by utilizing a multi-scale edge detection algorithm; processing the extracted edges and dividing the extracted edges into oil drop adhering images and oil drop non-adhering images; dividing the oil drop adhering images by utilizing a watershed algorithm, and then performing circle fitting operation; directly performing circle fitting operation on the oil drop non-adhering image; integrating two fitting results, so as to obtain an oil drop edge extraction result chart; and calculating the diameter D of each oil drop. By using the measuring method of the oil drop sizes, which is provided by the invention, the size of each oil drop can be accurately measured on the basis of obtaining a high-definition oil-water two-phase flow image.