CN110907469A

CN110907469A - Circuit board welding detection device using machine vision

Info

Publication number: CN110907469A
Application number: CN201911309506.XA
Authority: CN
Inventors: 胡小春
Original assignee: Hunan Smart Automation Technology Co Ltd
Current assignee: Hunan Smart Automation Technology Co Ltd
Priority date: 2019-12-18
Filing date: 2019-12-18
Publication date: 2020-03-24

Abstract

The invention discloses a circuit board welding detection device using machine vision, which comprises a two-dimensional electric rotating platform for mounting a circuit board to be detected, a group of vision cameras containing light sources and a computer platform for processing images acquired by the vision cameras. The two-dimensional electric rotating platform controls the pose of the circuit board to be detected so as to realize image acquisition of multiple incidence angles of a square welding spot, the visual camera comprising the light source has the light source emergent direction parallel to the lens direction of the camera and is used for acquiring the circuit board image, and the computer platform for processing the image acquired by the visual camera is used for identifying the welding spot by machine vision and judging whether the welding spot is normal or not.

Description

Circuit board welding detection device using machine vision

Technical Field

The invention relates to the field of machine vision and the technical field of circuit board welding detection, in particular to a circuit board welding detection device using machine vision.

Background

The circuit board is an important component of modern industrial products, product detection is one of important processes in the manufacturing process of the circuit board, the patch is one of basic units for realizing functions of the modern circuit board, the development trend of the current patch is precision and miniaturization, and an accurate and high-efficiency detection method of a Printed Circuit Board (PCB) is needed;

the picture that ordinary camera was taken is two-dimensional image, but the superficial layer nature of the soldering tin at paster both ends is three-dimensional detection target, and modern PCB inspection industry widely adopts RGB three-colour light source irradiation paster, to the different colors of soldering tin reflection of different gradients, thereby reflected the soldering tin surface nature, printed circuit board paster has slight difference in characteristics such as paster position, angle, solder joint shape, but relative to the size of paster, the influence of this slight difference is not negligible, and traditional image identification method, such as image contrast method, too sensitive to slight difference, easily cause the erroneous judgement, but other morphological image identification algorithm, such as wavelet transform, hough transform, have the calculated amount big shortcoming of influence by background interference, etc., the testing result can not satisfy the requirement completely.

Disclosure of Invention

The invention provides a circuit board welding detection device using machine vision, which can effectively solve the problem that the detection result provided by the background technology can not completely meet the requirement.

In order to achieve the purpose, the invention provides the following technical scheme: a circuit board welding detection device using machine vision comprises a two-dimensional electric rotating platform for mounting a circuit board to be detected, a group of vision cameras containing light sources and a computer platform for processing images acquired by the vision cameras;

the two-dimensional electric rotating platform controls the pose of the circuit board to be detected so as to realize the image acquisition of the square multi-incidence angle of the welding spot;

the emergent direction of the light source of the visual camera containing the light source is parallel to the direction of a lens of the camera and is used for collecting the circuit board image;

and the computer platform for processing the image acquired by the vision camera is used for identifying the welding spot by machine vision and judging whether the welding spot is normal or not.

A detection method of a circuit board welding detection device using machine vision comprises the following steps: images are shot at the square positions of the welding pins and at multiple discrete included angles (smaller than 45 degrees) with the plane of the circuit board, and whether normal welding is carried out is judged by comparing identification parameters of elements in a normal motherboard at the same angle through geometric characteristics such as the area size of a welding spot identified by the images.

Preferably, during the visual recognition, a threshold value binarization method is used for image segmentation, and the welding spot area is a bright spot area.

Preferably, when the welding spot is judged to be abnormal, the area ratio of each angle is used to compare with the size of the area.

Compared with the prior art, the invention has the beneficial effects that: the invention has scientific and reasonable structure, safe and convenient use, simplifies the algorithm of machine vision identification, greatly improves the detection reliability, such as the defects of missing parts, wrong parts, component deviation, tombstone and welding spots, uses the template patch, calculates the threshold value according to the algorithm program, and can realize the detection of the patch defect.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention.

In the drawings:

FIG. 1 is a schematic view of a measurement angle of a solder joint of a surface mount component according to the present invention;

FIG. 2 is a schematic diagram of the detection structure of the present invention.

Detailed Description

The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it will be understood that they are described herein for the purpose of illustration and explanation and not limitation.

Example 1: as shown in fig. 2, the present invention provides a technical solution, a circuit board welding detection apparatus using machine vision, comprising a two-dimensional electric rotating platform for mounting a circuit board to be detected, a group of vision cameras including a light source, and a computer platform for processing images acquired by the vision cameras;

the visual camera comprises a light source, wherein the emergent direction of the light source is parallel to the direction of a lens of the camera and is used for collecting the circuit board image;

and the computer platform is used for processing the images acquired by the vision camera and is used for identifying the welding points through machine vision and judging whether the welding points are normal or not.

Example 2: as shown in fig. 1, the present invention provides a technical solution, a method for detecting a circuit board soldering detecting apparatus using machine vision, comprising the steps of: images are shot at the square positions of the welding pins and at multiple discrete included angles (smaller than 45 degrees) with the plane of the circuit board, and the identification parameters of all angles of elements in a normal motherboard are compared with the identification parameters of all angles of the elements in the normal motherboard through the geometric characteristics such as the area size of the welding points identified by the images, so that whether normal welding is carried out is judged.

In the invention, the included angle less than 45 degrees is selected because the reflection of the welding spot is quite obvious at a small angle, thereby generating a good contrast effect; on the contrary, at a large angle, the effect is poor. Further, during the visual identification, a threshold value binarization method is used for image segmentation, and a welding spot area is a bright spot area.

Further, when judging whether the welding spot is abnormal, comparing by using the area ratio of each angle and the size of the area.

Further, the direction indicated by the arrow is defined as the positive direction of the patch element, the solder feet of the patch element generally have four directions of 0,90,180 and 270, and other angles are included, so that image acquisition and identification are carried out from all positive directions of the patch elements when the circuit board is measured;

for pin pads, any orientation is square due to rotational symmetry of the pad, and can be measured in 4 orientations, typically 0,90,180, 270.

Furthermore, the lens axis direction of the vision camera is parallel to the light source incidence direction, the image received by the vision camera is light source reflection light, and the reflection effect is optimal when the light source incidence direction and the circuit board form a small angle (generally 30 degrees, and 45-15 degrees can be selected).

Further, four discrete angles of 15,25,35 and 45 are selected as measuring angles, the welding point area recognized by machine vision at each angle is different, if the welding point is abnormal, the welding point recognized by machine vision is too different from the sample plate, and thus the abnormality is judged. And a plurality of included angles are selected for comparison, so that the misjudgment probability is very small.

Furthermore, in order to realize the test of a plurality of directions and included angles of the circuit board, a two-dimensional numerical control electric rotating platform is used for realizing the pose control;

the computer platform is matched with the visual camera and is used for processing images acquired by the visual camera, and the shooting and the visual recognition of the camera are not carried out by taking one welding point as a unit, but by taking the whole plate or one area of the plate as a unit, the whole plate or one area of the plate comprises a plurality of welding points.

Further, since the welding spots have strong light reflectivity, it is preferable to perform image segmentation using a threshold binarization method during image processing, and the welding spot regions are bright spot regions. And simultaneously, the method preferably supports manual elimination of interfered non-welding spots on visual software.

The welding element may have position deviation, so that the area size of a welding spot area is deviated, and whether welding is successful or not is difficult to judge by the area size in the identification process, but the area ratio of each angle can be used for comparison by combining the area size and other geometric parameters, and the comparison by combining the area ratio of each angle with the area size of each angle is proved to be sufficient for 99% detection occasions by practice.

Further, as a supplementary explanation of the above scheme combining the comparison and identification of the area ratio of each angle and the area size of each angle, assuming that the area of a normal welding spot area at four angles {15,25,35,45} is {35,30,25,18} and the rule is decreasing, and the area obtained by measuring the same welding spot on another board is {25,27,35,24}, it is theoretically possible to allow a large deviation of the area at each angle, but the area at an angle is proportional to the area at a normal welding spot, for example, {25,21,17,12} is normal, and {25,27,35,24} is found by the area ratio, and four groups of data are not proportional, it can be determined as abnormal, and the error of the set ratio is generally in the range of 0.9 to 1.1.

Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that changes may be made in the embodiments and/or equivalents thereof without departing from the spirit and scope of the invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A circuit board welding detection device using machine vision comprises a two-dimensional electric rotating platform for mounting a circuit board to be detected, a group of vision cameras containing light sources and a computer platform for processing images acquired by the vision cameras;

the computer platform for processing the image acquired by the visual camera is used for identifying the welding spot by machine vision and judging whether the welding spot is normal or not;

the method is characterized in that:

images are shot at the square positions of the welding pins and at multiple discrete included angles (smaller than 45 degrees) with the plane of the circuit board, and whether normal welding is carried out is judged by comparing identification parameters of elements in a normal motherboard at the same angle through geometric characteristics such as the area size of a welding spot identified by the images.

2. The method as claimed in claim 1, wherein the visual recognition is performed by image segmentation using a threshold binarization method, and the solder joint area is a bright spot area.

3. The method as claimed in claim 2, wherein the determination of the solder joints is made by comparing the area ratio of each angle with the area size.