CN113052797B - BGA solder ball three-dimensional detection method based on depth image processing - Google Patents
BGA solder ball three-dimensional detection method based on depth image processing Download PDFInfo
- Publication number
- CN113052797B CN113052797B CN202110249577.6A CN202110249577A CN113052797B CN 113052797 B CN113052797 B CN 113052797B CN 202110249577 A CN202110249577 A CN 202110249577A CN 113052797 B CN113052797 B CN 113052797B
- Authority
- CN
- China
- Prior art keywords
- solder ball
- dimensional
- depth image
- point
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 229910000679 solder Inorganic materials 0.000 title claims abstract description 121
- 238000001514 detection method Methods 0.000 title claims abstract description 44
- 238000012545 processing Methods 0.000 title claims abstract description 33
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000004458 analytical method Methods 0.000 claims abstract description 17
- ATJFFYVFTNAWJD-UHFFFAOYSA-N Tin Chemical compound [Sn] ATJFFYVFTNAWJD-UHFFFAOYSA-N 0.000 claims abstract description 16
- 230000000877 morphologic effect Effects 0.000 claims abstract description 14
- 238000006243 chemical reaction Methods 0.000 claims abstract description 6
- UDHXJZHVNHGCEC-UHFFFAOYSA-N Chlorophacinone Chemical compound C1=CC(Cl)=CC=C1C(C=1C=CC=CC=1)C(=O)C1C(=O)C2=CC=CC=C2C1=O UDHXJZHVNHGCEC-UHFFFAOYSA-N 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000011895 specific detection Methods 0.000 claims description 3
- 238000007689 inspection Methods 0.000 claims 4
- 238000005260 corrosion Methods 0.000 claims 1
- 230000007797 corrosion Effects 0.000 claims 1
- 238000010606 normalization Methods 0.000 claims 1
- 238000005516 engineering process Methods 0.000 abstract description 17
- 230000007547 defect Effects 0.000 description 7
- 238000002474 experimental method Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000005259 measurement Methods 0.000 description 5
- 230000011218 segmentation Effects 0.000 description 3
- 238000003466 welding Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000005530 etching Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000004806 packaging method and process Methods 0.000 description 2
- 238000011897 real-time detection Methods 0.000 description 2
- HPTJABJPZMULFH-UHFFFAOYSA-N 12-[(Cyclohexylcarbamoyl)amino]dodecanoic acid Chemical compound OC(=O)CCCCCCCCCCCNC(=O)NC1CCCCC1 HPTJABJPZMULFH-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20036—Morphological image processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30148—Semiconductor; IC; Wafer
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Length Measuring Devices By Optical Means (AREA)
- Electric Connection Of Electric Components To Printed Circuits (AREA)
Abstract
The three-dimensional BGA solder ball detection method based on depth image processing comprises the steps of converting an acquired chip point cloud into a depth image, fitting a base plane of a chip and marking each solder ball area by using a Blob analysis algorithm; extracting a three-dimensional coordinate point corresponding to a two-dimensional center point of the tin ball area and an 8 neighborhood pixel of the tin ball area, and carrying out average assignment on the three-dimensional coordinate point to a three-dimensional point at the top of the tin ball; when marking a solder ball area, a morphological method is combined to enable the solder ball to approach to circularization so as to improve the robustness of an algorithm; and finally, calculating the distance from the vertex to the plane to obtain the height and coplanarity of the solder balls. According to the BGA solder ball three-dimensional detection method based on depth image processing, the machine vision technology is utilized, after the depth map conversion is carried out on the acquired chip point cloud by utilizing the point cloud and image processing technology, the solder ball three-dimensional detection index can be obtained efficiently, and the precision meets the detection requirement.
Description
Technical Field
The invention relates to the technical field of point cloud processing and image processing; in particular to a BGA solder ball three-dimensional defect detection method based on depth image processing.
Background
With the continuous development of chip industry chains in China for domestic and foreign enterprises, chip packaging detection technology is rapidly developed. Ball Grid Array (BGA) chips are one of many integrated chips, and the packaging quality of the Ball-mounting process directly affects the performance and reliability of the chip. The characteristics of small diameter of the solder ball and high precision of the detection index lead the common contact type measurement technology not to meet the requirements. The detection technology applied to the BGA chip at present mainly comprises the following steps: solder ball detection based on X-ray, solder ball detection based on planar image processing technology, solder ball surface defect detection based on machine learning, and the like. The detection technology based on X-rays is nondestructive detection, but the detection efficiency is not high and depends on subjective judgment of detection personnel; the detection technology based on plane image processing mainly aims at the surface defects of the solder ball, such as the radius, the distance, the offset and the like of the solder ball, so that the three-dimensional index of the solder ball cannot be measured, and the surface of the solder ball is smooth and easy to reflect light, so that the requirements on a lighting system are very strict during shooting; the detection technology based on machine learning is mainly aimed at the surface defects of the solder balls, such as empty welding, over welding, lack welding, solder ball adhesion and the like. The three-dimensional size of the BGA solder ball is measured by a learner by adopting a CCD, but the modeling of the method is complex, and the requirement on hardware facilities is extremely high. Thus, no published literature report has been reported to find detection from the chip point cloud. Although the point cloud has the characteristic of high precision, the high time consumption of the processing method can not meet the real-time detection requirement of industry.
Disclosure of Invention
The invention aims to provide a three-dimensional detection method for the tin balls of the BGA chip, which has good robustness and high precision and high speed. The method is based on the principle that the obtained BGA chip point cloud is converted into a depth image, and image processing technologies such as Blob analysis and the like are introduced for testing, so that the height and coplanarity of each solder ball of the chip can be obtained rapidly. The algorithm running speed can also reach the real-time detection requirement.
In order to achieve the above purpose, the technical scheme of the invention is as follows:
firstly, converting an acquired chip point cloud into a depth image, fitting a base plane of the chip, marking each solder ball area by using a Blob analysis technology, extracting a three-dimensional coordinate point corresponding to a two-dimensional center point of the solder ball area and an 8 neighborhood pixel of the solder ball area, and carrying out average assignment on the three-dimensional coordinate point to the top of the solder ball. To improve the robustness of the algorithm, the tin ball is approximately rounded by combining a morphological method when the tin ball region is marked. And finally, calculating the distance from the vertex to the plane to obtain the height and coplanarity of the solder balls. The method comprises the following specific steps:
step1: and acquiring BGA chip point cloud data. And according to the specific detection environment, adjusting parameters such as light intensity, exposure time and the like of the binocular line scanning camera to obtain complete 3D point cloud data.
Step2: and performing depth map conversion on the acquired chip point cloud. According to the distance between each point of the point cloud, the size of a projection image and the distance between pixels are preset, 3D point cloud data can be converted into a depth image (also called a distance image) by using a method of projecting along the Z direction and normalizing in the XY direction, namely, each pixel in the depth image corresponds to each three-dimensional point in the point cloud one by one, the bit depth of the depth image is 16 bits, and the gray value of each pixel represents the distance.
Step3: the chip base area in the depth image is extracted and used for plane fitting. For the depth image, through the relative height, area and shape characteristic information of a preset BGA chip point cloud base, firstly extracting 3D point data of the base position and envelope of the BGA, and fitting a chip base plane by using a least square method or a Ranac sampling consistency algorithm to obtain a plane equation.
Step4: and extracting a chip tin ball area in the depth image. Extracting and marking the position and the enveloped 3D point data of each solder ball by utilizing the improved Blob analysis method through the preset height, area and shape characteristic information of the solder ball area relative to the base, and obtaining the two-dimensional center coordinates of each area, namelyWherein x is c 、y c For the two-dimensional center coordinates of each solder ball area, x i 、y i And N is the total number of pixels for all pixel point coordinates in each solder ball area.
The improved Blob analysis algorithm mainly comprises:
(1) And (5) region binarization. The binarization operation is a commonly used segmentation method in machine vision, and is characterized in that the threshold value is selected, and the main method is divided into methods based on edge gray scale, histogram clustering and fixed value selection. According to the invention, for the depth image, the binarization operation is performed by setting a height constant value of the solder ball area relative to the Z direction of the base as a threshold value T. Inputting 16-bit image g (x, y) to obtain 8-bit binary image g b (x,y)：
(2) Morphological treatment. By setting upAfter the binarization operation is carried out on the height threshold value of the solder ball, the target is inevitably divided into discontinuous areas, noise points among the areas are removed and holes of the areas are filled in order to reduce calculation errors, and the morphological processing is very important. First of all use is made of circular structural elements M 1 Performing a closing operation on the binarized image G to obtain an image G 1 The definition is as follows
Then for image G 1 By circular structure M 2 Performing open operation to remove isolated noise points to obtain an image G 1 To ensure the integrity of each solder ball area, the definition is as follows
Wherein, is a closed operation;is an open operation; />Is an expansion operation; />Is an etching operation.
(3) Blob analysis. The image is divided into a background area and a target area through binarization and morphological processing, and for a binary image, the Blob analysis actually adopts a 4 or 8 neighborhood method to cluster pixels with the same or similar pixel values, so that coordinate values and total pixel numbers of all pixels in various types can be obtained. After two-dimensional center points of the solder ball areas in the image and 8 neighborhood pixel coordinates of the two-dimensional center points are calculated, three-dimensional coordinate points corresponding to the 9 points are extracted to be averaged together, and the three-dimensional points are assigned to the top three-dimensional points of the solder balls, so that errors caused by inaccuracy in center point selection are reduced, and the robustness of an algorithm is improved.
Step5: and sequencing the solder balls in sequence according to the maximum and minimum values of the horizontal and vertical coordinates of the center of each solder ball area. Calculating the distance between the point and the plane of the base by using the three-dimensional coordinate point at the top of the solder ball to obtain the height of the solder ball; the average value of the heights of the solder balls is obtained, and the position deviation of the solder balls can be obtained by using a difference method to obtain the coplanarity. Let the top coordinate point of each solder ball be p (x) i ,y i ,z i ) I is {1,2, … m }, the height and coplanarity of each solder ball are calculated as follows
H in the above i The height of each solder ball is equal to the height of each solder ball; c (C) i Coplanarity of each solder ball; m is the total number of tin balls; A. b, C, D are equation coefficients for the base plane fit.
Compared with the prior art, the invention has the beneficial effects that:
according to the BGA solder ball three-dimensional detection method based on depth image processing, the machine vision technology is utilized, after the depth map conversion is carried out on the collected chip point cloud by utilizing the point cloud and image processing technology, the solder ball three-dimensional detection index can be obtained quickly and efficiently, and the precision also meets the detection requirement.
Drawings
FIG. 1 is a schematic diagram of three-dimensional detection indexes of solder balls of a BGA chip;
FIG. 2 is a flow chart of the overall scheme of the present invention;
FIG. 3 point cloud and converted depth map;
FIG. 4 binarization processing;
FIG. 5 morphological treatment;
fig. 6 is a close-up of the tin ball midpoint;
FIG. 7 shows the detection results.
Detailed Description
The 3D size detection indexes of the BGA chip are mainly divided into solder ball height (height), solder ball coplanarity (coplanarity), chip bow (package) and chip base height (package height). The invention mainly aims at tin balls to realize the detection of the height and coplanarity. The detection index is shown in fig. 1, wherein the height is the distance from the top of the solder ball to the plane of the chip base, and the coplanarity takes the height offset of the solder ball as a measurement index.
With the continuous development of the chip industry and the continuous perfection of the chip detection industry chain, the traditional manual detection efficiency is low, the cost of the laser detection method is too high, the defects of high detection efficiency and high detection index accuracy of a detection system are overcome in order to overcome the defects of low efficiency and no pertinence in the prior art, and the detection requirements of the three-dimensional index of the BGA chip solder ball in FIG. 1 can be met by an algorithm. After the depth map conversion is carried out on the acquired chip point cloud by utilizing the point cloud and image processing technology, the method can efficiently obtain the three-dimensional detection index of the solder ball in fig. 1, and the precision meets the detection requirement. The technical scheme of the invention is further described in detail below by combining examples.
Example 1
As shown in FIG. 2, in the BGA solder ball three-dimensional defect detection method based on depth image processing, firstly, the acquired chip point cloud is converted into a depth image, the base plane of the chip is fitted, each solder ball area is marked by using a Blob analysis technology, then, the three-dimensional coordinate points of the two-dimensional center point of the solder ball area and the 8 neighborhood pixels are extracted, and the average value is assigned to the three-dimensional point at the top of the solder ball. To improve the robustness of the algorithm, the tin ball is approximately rounded by combining a morphological method when the tin ball region is marked. And finally, calculating the distance from the vertex to the plane to obtain the height and coplanarity of the solder balls.
In the embodiment, BGA chip point cloud data is obtained through a binocular 3D line scanning camera with the model of Chromasens-3DPIXA, and the number of the point clouds is 2273758 by utilizing a CloudCompare software platform for display. One BGA chip; one camera fixing bracket; the CPU model of the computer hardware configuration is Intel (R) Core (TM) i7-8700, the model of the GPU is NVIDIA GeForce GTX 1660 graphics card, and the memory size is 16.0GB. Algorithms were developed using Visual Studio 2017+opencv library, etc. In terms of speed, based on depth maps and parallel processing mechanisms, openMP and multithreading technologies are used on a CPU, and a high-performance CUDA programming technology is used on a GPU, so that a plurality of BGA chip point clouds can be processed simultaneously.
The method comprises the following specific steps:
step1: and acquiring BGA chip point cloud data. And according to the specific detection environment, adjusting parameters such as light intensity, exposure time and the like of the binocular line scanning camera to obtain complete 3D point cloud data.
Step2: and performing depth map conversion on the acquired chip point cloud. According to the distance between each point of the point cloud, the size of a projection image and the distance between pixels are preset, 3D point cloud data can be converted into a depth image (also called a distance image) by using a method of projecting along the Z direction and normalizing in the XY direction, namely, each pixel in the depth image corresponds to each three-dimensional point in the point cloud one by one, the bit depth of the depth image is 16 bits, and the gray value of each pixel represents the distance.
Step3: the chip base area in the depth image is extracted and used for plane fitting. For the depth image, through the relative height, area and shape characteristic information of a preset BGA chip point cloud base, firstly extracting 3D point data of the base position and envelope of the BGA, and fitting a chip base plane by using a least square method or a Ranac sampling consistency algorithm to obtain a plane equation.
Step4: and extracting a chip tin ball area in the depth image. Extracting and marking the position and the enveloped 3D point data of each solder ball by utilizing the improved Blob analysis method through the preset height, area and shape characteristic information of the solder ball area relative to the base, and obtaining the two-dimensional center coordinates of each area, namelyWherein x is c 、y c For the two-dimensional center coordinates of each solder ball area, x i 、y i And N is the total number of pixels for all pixel point coordinates in each solder ball area.
The improved Blob analysis algorithm mainly comprises:
(1) And (5) region binarization. The binarization operation is a commonly used segmentation method in machine vision, and is characterized in that the threshold value is selected, and the main method is divided into methods based on edge gray scale, histogram clustering and fixed value selection. According to the invention, for the depth image, the binarization operation is performed by setting a height constant value of the solder ball area relative to the Z direction of the base as a threshold value T. Inputting 16-bit image g (x, y) to obtain 8-bit binary image g b (x,y)：
(2) Morphological treatment. After binarization operation is performed through the set height threshold value, the target is inevitably divided into discontinuous areas, noise points among the areas are removed and cavities of the areas are filled, so that the tin ball areas are close to circularization, and morphological processing is very important. First of all use is made of circular structural elements M 1 Performing a closing operation on the binarized image G to obtain an image G 1 The definition is as follows
Then for image G 1 By circular structure M 2 Performing open operation to remove isolated noise points to obtain an image G 1 To ensure the integrity of each solder ball area, the definition is as follows
Wherein, is a closed operation;is an open operation; />Is an expansion operation; />Is an etching operation.
(3) Blob analysis. The image is divided into a background area and a target area through binarization and morphological processing, and for a binary image, the Blob analysis actually adopts a 4 or 8 neighborhood method to cluster pixels with the same or similar pixel values, so that coordinate values and total pixel numbers of all pixels in various types can be obtained. After two-dimensional center points of the solder ball areas in the image and 8 neighborhood pixel coordinates of the two-dimensional center points are calculated, three-dimensional coordinate points corresponding to the 9 points are extracted to be averaged together, and the three-dimensional points are assigned to the top three-dimensional points of the solder balls, so that errors caused by inaccuracy in center point selection are reduced, and the robustness of an algorithm is improved.
Step5: and sequencing the solder balls in sequence according to the maximum and minimum values of the horizontal and vertical coordinates of the center of each solder ball area. Calculating the distance between the point and the plane of the base by using the three-dimensional coordinate point at the top of the solder ball to obtain the height of the solder ball; the average value of the heights of the solder balls is obtained, and the position deviation of the solder balls can be obtained by using a difference method to obtain the coplanarity. Let the top coordinate point of each solder ball be p (x) i ,y i ,z i ) I is {1,2, … m }, the height and coplanarity of each solder ball are calculated as follows
H in the above i The height of each solder ball is equal to the height of each solder ball; c (C) i Coplanarity of each solder ball; m is the total number of tin balls; A. b, C, D are equation coefficients for the base plane fit.
Fig. 3 is an acquired point cloud and a converted depth image, wherein the left image is a chip 3D point cloud, and the right image is a depth image. Fig. 4 is a diagram showing the result of binarization processing, and the right diagram is a close-up diagram showing the left diagram, and it is apparent that there is a discontinuity in the binarized area of each solder ball. Fig. 5 is a result of processing the binary image by using a morphological processing technique, wherein the left image is morphological processing, the right image is a partial area close-up, and it is obvious that each solder ball area is more complete and more approximate to circularization, and no isolated noise point exists in the interval. After Blob analysis, the corresponding three-dimensional coordinates are extracted by utilizing the central coordinates of each solder ball region and 8 neighborhood pixels thereof, the three-dimensional coordinates of the top of the solder ball are obtained, and the solder balls are marked in the depth image. Fig. 6 is a close-up view of a center point mark of a portion of a solder ball area in a depth image.
After the three-dimensional points on the tops of the solder balls are calculated, the heights and coplanarity of the solder balls can be calculated by using the formula, and the result is drawn into fig. 7, wherein the abscissa of the figure is 552 solder balls in a chip, and the ordinate of the figure is the calculation result. From a runtime point of view, the detection algorithm for this chip of the present invention takes 986.34ms in total. According to the invention, the obtained 3D point cloud of the chip is directly processed, the base point cloud is removed based on a PCL (Point Cloud Library) algorithm library, each solder ball is independently segmented out and calculated by utilizing an European cluster segmentation algorithm, and the algorithm consumes about 14.392s. Therefore, the time consumption of the algorithm is far less than that of direct processing of the point cloud, and the three-dimensional detection method provided by the invention can be effectively used for real-time three-dimensional detection of the solder balls of the BGA chip.
Five experiments are carried out on the same chip under the same environment by using the method, five groups of point cloud data are collected in each experiment, and the height data of each solder ball of the chip are obtained as true values by using more precise equipment (the measurement precision is +/-2 mu m). Calculating average value E of absolute error of solder ball height of each group and solder ball height true value of the chip, taking maximum value of E in each experiment as measurement precision of each experiment, and calculating as follows
Wherein h is i 、h′ i Respectively measuring and truth value of each solder ball; n is the total number of solder balls of the chip; as can be seen from Table 1 below, the E maximum per experiment is not more than 0.01mm (industryThe accuracy of the internal detection standard is 0.015 mm), the high-accuracy requirement of measurement is met, and the feasibility and the effectiveness of the method for obtaining the height and the coplanarity of the solder balls of the BGA chip by utilizing depth image processing are verified.
TABLE 1 maximum value of E (mm) in each experiment
Claims (5)
1. The three-dimensional BGA solder ball detection method based on depth image processing is characterized by comprising the following steps of:
s1: acquiring BGA chip point cloud data;
s2: performing depth map conversion on the acquired chip point cloud;
s3: extracting a chip base region in the depth image and performing plane fitting on the chip base region;
s4: extracting and marking the position and the enveloped 3D point data of each solder ball by using a Blob analysis algorithm through the preset height, area and shape characteristic information of the solder ball area relative to the base, and obtaining the two-dimensional center coordinates of each area, namelyWherein x is c 、y c Is the two-dimensional center coordinate of the solder ball area, x i 、y i The coordinates of all pixel points in each solder ball area are given, and N is the total number of pixels;
the Blob analysis algorithm includes:
1) Region binarization:
for the depth image, performing binarization operation by setting a height constant value of the solder ball area relative to the Z direction of the base as a threshold T; inputting 16-bit image g (x, y) to obtain 8-bit binary image g b (x，y)：
2) Morphological treatment:
first of all use is made of circular structural elements M 1 Performing a closing operation on the binarized image G to obtain an image G 1 The definition is as follows
Then for image G 1 By circular structure M 2 Performing open operation to remove isolated noise points to obtain an image G 1 To ensure the integrity of each solder ball area, the definition is as follows
Wherein, is a closed operation;is an open operation; />Is an expansion operation; />Is a corrosion operation;
3) Blob analysis:
dividing an image into a background area and a target area through binarization and morphological processing, and clustering pixels with the same or similar pixel values by adopting a 4 or 8 neighborhood method for Blob analysis on the binary image to obtain coordinate values and total pixel numbers of all pixels in various types; after two-dimensional center points of each solder ball area in an image and 8 neighborhood pixel coordinates of the two-dimensional center points are calculated, three-dimensional coordinate points corresponding to the 9 points are extracted to be averaged together, and the three-dimensional points are assigned to the top three-dimensional points of the solder balls, so that errors caused by inaccuracy in center point selection are reduced, and the robustness of an algorithm is improved;
s5: sequencing the solder balls in sequence according to the maximum and minimum values of the central horizontal and vertical coordinates of the solder ball areas; calculating the distance between the point and the plane of the base by using the three-dimensional coordinate point at the top of the solder ball to obtain the height of the solder ball; the average value of the heights of the solder balls is obtained, and the position deviation of the solder balls can be obtained by using a difference method to obtain the coplanarity.
2. The three-dimensional inspection method for BGA solder balls based on depth image processing according to claim 1, wherein S1 comprises: and according to the specific detection environment condition, adjusting the light intensity and the exposure time of the binocular line scanning camera to obtain complete 3D point cloud data.
3. The three-dimensional inspection method for BGA solder balls based on depth image processing according to claim 1, wherein S2 comprises: according to the distance between each point of the point cloud, the size and the pixel distance of a projection image are preset, and 3D point cloud data are converted into a depth image by using a method of projection along the Z direction and normalization in the XY direction; that is, each pixel in the depth image corresponds to each three-dimensional point in the point cloud one by one, the bit depth of the depth image is 16 bits, and the gray value of each pixel represents the distance.
4. The three-dimensional inspection method for BGA solder balls based on depth image processing according to claim 1, wherein S3 comprises: for the depth image, through the relative height, area and shape characteristic information of a preset BGA chip point cloud base, firstly extracting 3D point data of the base position and envelope of the BGA, and fitting a chip base plane by using a least square method or a Ranac sampling consistency algorithm to obtain a plane equation.
5. The three-dimensional inspection method for BGA solder balls based on depth image processing according to claim 1, wherein S5 comprises: let the top coordinate point of each solder ball be p (x) i ，y i ，z i ) I is {1,2, … m }, the height and coplanarity of each solder ball are calculated as follows
Wherein h is i The height of each solder ball is equal to the height of each solder ball; c (C) i Coplanarity of each solder ball; m is the total number of tin balls; A. b, C, D are equation coefficients for the base plane fit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110249577.6A CN113052797B (en) | 2021-03-08 | 2021-03-08 | BGA solder ball three-dimensional detection method based on depth image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110249577.6A CN113052797B (en) | 2021-03-08 | 2021-03-08 | BGA solder ball three-dimensional detection method based on depth image processing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113052797A CN113052797A (en) | 2021-06-29 |
| CN113052797B true CN113052797B (en) | 2024-01-05 |
Family
ID=76510583
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110249577.6A Active CN113052797B (en) | 2021-03-08 | 2021-03-08 | BGA solder ball three-dimensional detection method based on depth image processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113052797B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114267606B (en) * | 2022-03-01 | 2022-06-21 | 武汉精立电子技术有限公司 | Wafer height detection method and device |
| CN115908338B (en) * | 2022-11-29 | 2023-08-29 | 北京数字绿土科技股份有限公司 | Three-dimensional point cloud-based electric power facility and intrusion object distance detection method and system |
| CN116758088B (en) * | 2023-08-22 | 2023-12-22 | 深圳市立可自动化设备有限公司 | Chip detection method for Ball Grid Array (BGA) ball mounting and ball mounting system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1841008A (en) * | 2005-04-01 | 2006-10-04 | 鸿富锦精密工业（深圳）有限公司 | Verification method for scanning precision of laser measuring machine |
| CN106017356A (en) * | 2016-08-01 | 2016-10-12 | 南京理工大学 | Three-dimensional microscopic surface profile measuring device based on Greenough-type stereomicroscope and a method of device |
| CN107077735A (en) * | 2014-10-28 | 2017-08-18 | 惠普发展公司，有限责任合伙企业 | Three dimensional object is recognized |
| CN108317979A (en) * | 2017-12-11 | 2018-07-24 | 天津津航计算技术研究所 | A method of measuring chip bga soldered ball coplane degree |
| CN111223053A (en) * | 2019-11-18 | 2020-06-02 | 北京邮电大学 | Data enhancement method based on depth image |
| CN112082477A (en) * | 2020-09-01 | 2020-12-15 | 中国科学技术大学 | Universal tool microscope three-dimensional measuring device and method based on structured light |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8442304B2 (en) * | 2008-12-29 | 2013-05-14 | Cognex Corporation | System and method for three-dimensional alignment of objects using machine vision |
-
2021
- 2021-03-08 CN CN202110249577.6A patent/CN113052797B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1841008A (en) * | 2005-04-01 | 2006-10-04 | 鸿富锦精密工业（深圳）有限公司 | Verification method for scanning precision of laser measuring machine |
| CN107077735A (en) * | 2014-10-28 | 2017-08-18 | 惠普发展公司，有限责任合伙企业 | Three dimensional object is recognized |
| CN106017356A (en) * | 2016-08-01 | 2016-10-12 | 南京理工大学 | Three-dimensional microscopic surface profile measuring device based on Greenough-type stereomicroscope and a method of device |
| CN108317979A (en) * | 2017-12-11 | 2018-07-24 | 天津津航计算技术研究所 | A method of measuring chip bga soldered ball coplane degree |
| CN111223053A (en) * | 2019-11-18 | 2020-06-02 | 北京邮电大学 | Data enhancement method based on depth image |
| CN112082477A (en) * | 2020-09-01 | 2020-12-15 | 中国科学技术大学 | Universal tool microscope three-dimensional measuring device and method based on structured light |
Non-Patent Citations (5)
| Title |
|---|
| BGA焊点检测技术研究;胥路平 等;计量与测试技术(第05期);全文 * |
| 基于单幅图像的集成电路引脚共面性检测方法;吴福培 等;光学学报(第01期);全文 * |
| 基于激光三角法的接插件引脚共面度检测;张远 等;传感器与微系统(第09期);全文 * |
| 基于高精度轮廓线激光的芯片引脚共面度视觉检测系统;宋丽梅 等;光学技术(第06期);全文 * |
| 球栅阵列芯片锡球的三维检测方法研究;梁天为 等;传感器与微系统;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113052797A (en) | 2021-06-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113052797B (en) | BGA solder ball three-dimensional detection method based on depth image processing | |
| US9704232B2 (en) | Stereo vision measurement system and method | |
| CN104897062B (en) | Visual measurement method and device for shape and position deviation of part non-coplanar parallel holes | |
| JP6598162B2 (en) | Visual identification method of multi-type BGA chip based on linear clustering | |
| CN105783786B (en) | Part chamfering measuring method and device based on structured light vision | |
| US20220198712A1 (en) | Method for adaptively detecting chessboard sub-pixel level corner points | |
| CN110415257B (en) | Gas-liquid two-phase flow overlapped bubble image segmentation method | |
| CN109900713B (en) | Camera-guided unmanned aerial vehicle wind power blade defect dynamic detection system and method thereof | |
| CN112651968A (en) | Wood board deformation and pit detection method based on depth information | |
| CN112348840A (en) | QFP chip pin defect discrimination method based on pixel region growth | |
| CN109211110A (en) | A kind of flat-type workpiece size detection system and method based on machine vision | |
| CN113074660B (en) | Surface shape measuring method for large-size transparent object | |
| CN115937203B (en) | Visual detection method, device, equipment and medium based on template matching | |
| CN115597494B (en) | Precision detection method and system for prefabricated part preformed hole based on point cloud | |
| CN103884294B (en) | The method and its device of a kind of infrared light measuring three-dimensional morphology of wide visual field | |
| Li et al. | Stereo vision based automated solder ball height and substrate coplanarity inspection | |
| CN104574367B (en) | Method for detecting and calculating geometric size of integrated chip pin based on machine vision | |
| CN110426395B (en) | Method and device for detecting surface of solar EL battery silicon wafer | |
| CN106018198B (en) | A kind of Inversion Calculation method of bubble diameter | |
| CN113570550B (en) | Gold wire bonding size detection method based on three-dimensional point cloud | |
| CN116579955B (en) | New energy battery cell weld reflection point denoising and point cloud complement method and system | |
| CN113705564A (en) | Pointer type instrument identification reading method | |
| CN104655041A (en) | Industrial part contour line multi-feature extracting method with additional constraint conditions | |
| CN116739986A (en) | Optical acoustic combination method for detecting typical defects of AUV in water conveyance tunnel | |
| CN104881652B (en) | A kind of line number automatic testing method based on corncob male features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |