CN108876851A - A kind of foil gauge image position method - Google Patents

Abstract

The invention is a method for positioning square strain gauges in a strain gauge template, the steps are as follows: (1) image preprocessing; (2) preliminary positioning of strain gauges; (3) extraction of the area where individual strain gauges are located (including non-strain gauges) part); (4) Individual strain gauges are distributed and extracted individual strain gauge areas; after the above steps, the individual strain gauge areas are obtained, and after positioning the strain gauges (usually the corner strain gauges of the strain gauge template), according to the position of the area and The standard distance between each strain gauge during processing is used to obtain the positions of all strain gauges in the strain gauge template. The beneficial effect of the present invention is that: the present invention can realize machine positioning of strain gauges instead of manual labor, and the speed is faster than that of manual labor, and the precision meets the positioning requirements of engineering actual grinding strain gauges.

Description

A Method for Image Positioning of Strain Gauge

technical field

The invention belongs to the fields of machine vision and industrial automation, and in particular relates to a strain gauge image positioning method.

Background technique

Resistance strain gauge is a widely used resistive sensor made of the principle of resistance effect. It is mainly used in the inspection of mechanical quantities, such as the detection of physical quantities such as gravity and pressure, and has high requirements for resistance accuracy. Conventional resistance strain gauges are all made by chemical corrosion, and the resistance value of the produced resistance strain gauges is low, which cannot meet the requirements for use. Therefore, the resistance strain gauge of the semi-finished product must be automatically adjusted before use.

At present, the automatic resistance adjustment of the resistance strain gauge is mainly carried out by manual operation. The operator holds a miniature DC motor, the output shaft end of the DC motor is wrapped with absorbent cotton, and the absorbent cotton is coated with abrasive paste. When trimming the resistance value, the absorbent cotton presses the grid wire part of the resistance strain gauge, and the DC motor drives the absorbent cotton to rotate to generate grinding force. Under the action of grinding force, the thickness of grid wire becomes thinner, so as to achieve the purpose of resistance trimming.

Obviously, this method has the following disadvantages: high labor intensity consumption; working hours, slow progress, low efficiency, which affects production capacity; there are different degrees of error, which affects the accuracy of the resistance value. Therefore, it is particularly urgent to develop a device that can automatically locate and correct the resistance value of the resistance strain gauge.

Contents of the invention

The technical problem to be solved by the present invention is to overcome the technical defects of slow manual grinding and sometimes insufficient precision, and provide a strain gauge image positioning method.

The technical solution of the present invention is, a kind of strain gage positioning method on template, described method comprises the following steps:

(1) Image preprocessing, preprocessing the collected strain gauge images, specifically including,

a. Image grayscale: the weighted average method processes the image to obtain the grayscale image of the original image;

b. Image binarization: use the threshold value obtained by the 1.25 times larger law to perform image binarization;

(2) Initial positioning of the strain gauges, extracting the connected domain and judging where is the connected area containing the strain gauges according to the area;

(3) Extraction of the area where individual strain gauges are located (including non-strain gauge parts), and the area where strain gauges are located at the edge corners of general screenshots;

(4) Individual strain gauges are processed separately and individual strain gauge areas are extracted. After the interception is completed, according to the number of pixels in the rows and columns, the gray levels of the rows and columns with a small number of pixels are all set to 0;

After the above steps, the strain gauge area is selected according to the area.

The beneficial effect of the present invention is that: the present invention can realize the machine positioning of the strain gauge instead of manual labor, and the speed is faster than that of manual labor, and the precision is sufficient, which meets the positioning requirements of engineering actual grinding strain gauges.

Description of drawings

Fig. 1 is a flow chart of strain gauge positioning in the present invention.

Detailed ways

Below, the present invention is described in detail as follows in conjunction with accompanying drawing:

There are many types of strain gauges, among which there are many square strain gauges, which can be delineated by the minimum circumscribed rotation rectangle in the image positioning. Because the parts outside the strain gauge are connected to the strain gauge, it is difficult to directly extract the strain gauge area. Therefore, on the basis of binarization, we propose a connected domain extraction method of decentralized processing to extract the strain gauge area and locate it.

Image preprocessing.

Image grayscale: image grayscale adopts the built-in function in the OPENCV toolkit to convert the RGB image into a grayscale image. Image binarization: In order to adapt to a certain range of lighting environment changes, the final result uses a multiple of one to two times the Dalaw threshold as the final dynamic threshold, and uses the dichotomy to calculate the hypothetical threshold between the Dalaw threshold and 255 , and after repeated experiments and comparisons, it is finally concluded that the threshold of 1.25 times the law is the best threshold for the experimental environment of the present invention.

Initial positioning of strain gauges.

Extract the connected domain and judge where is the connected area containing the strain gauge according to the area.

Extract the area where individual strain gauges are located (including non-strain gauge parts).

Generally, the screenshot shows the area where the strain gauges are located at the edge corners.

Individual strain gauges are dispersed and extracted for individual strain gauge regions.

(1) According to the area, a strain gauge pair is sequentially extracted each time;

(2) Perform rotation correction (for the convenience of decentralized processing);

(3) Scatter processing according to the number of pixels in each row and column (mainly remove the connection part of the strain gauge solder joint lead and the strain gauge block, that is, the separation of the non-strain gauge part and the strain gauge part), the number of pixels in a certain row or column is too large At least the gray value of the pixel in this row or column is set to 0;

(4) Select the connected domain of the strain gauge according to the area of the connected domain, and then rotate back to the original position according to the angle of the first step of rotation, and use the minimum circumscribed rotation rectangle of the connected domain of the strain gauge after rotation to locate the center position of the strain gauge.

So far, the central position of the strain gauge and the attitude of the strain gauge can be expressed by the angle between the long side and the horizontal axis. These two parameters can be obtained from the minimum circumscribed rotation rectangle of the connected domain of the strain gauge. The center coordinate represents its position, and the angle is its attitude.

The above are only specific examples of the present invention, and any equivalent transformation based on the method of the present invention falls within the protection scope of the present invention.

Claims (1)

1. a kind of foil gauge image position method, which is characterized in that the method handle includes the following steps：

（1）Image and processing, the foil gauge picture come to acquisition are pre-processed, are specifically included,

A. image gray processing：Weighted mean method handles image, obtains the grayscale image of original image;

B. image binaryzation：Image binaryzation is carried out using the threshold value two that 1.25 times big law obtains；

（2）Foil gauge just positions, and carries out connected domain and extracts and judge where be the connected region comprising foil gauge according to area；

（3）Extract (including unstrained part), general screenshot edge corner foil gauge location in individual foil gauge regions Domain；

（4）Individual foil gauge decentralized processings simultaneously extract individual strain panel regions, after the completion of interception, according to ranks pixel number, picture The few part ranks gray scale of prime number all sets 0；

By above-mentioned steps, strain panel region is selected further according to area.