TWI666426B - Liquid level detecting method - Google Patents

Abstract

A liquid surface height detection method is used to detect a liquid surface height of a solder wave rising from the inside of a solder bath. In this method, an image capturing device is first used to capture an image to be tested from the side of the solder bath, and a region of interest is selected from the image to be tested. An appropriate measurement process is selected from the complex measurement process based on the brightness data of an area of interest and an ambient light color. The selected measurement process is performed to obtain a pixel value representing the height of the liquid surface in the image to be measured, and the pixel value is converted into a real height measurement value at the wave welding site. This method uses non-contact real-time image analysis, which can adapt to the dynamic change of liquid surface under different brightness, and can obtain an accurate measurement result under the interference of colored light.

Description

Level detection method

The invention relates to a method for detecting the height of a liquid surface, and more particularly to a method for detecting the height of a liquid surface using a visualized image.

At present, the printed circuit board (PCB) of high-power circuits on the market still uses dual in-line package (DIP) components as the main design, and the welding of DIP components is still wave soldering (Wave Soldering).

FIG. 1 is a schematic diagram of a conventional wave welding system. The conventional wave soldering system 100 includes a conveying rail 110, a flux nozzle 120, a preheater 130, a solder tank 140, a fan 150, and a central controller 160. The printed circuit board 170 in which the DIP package element 172 is inserted is placed on the conveying rail 110, sprayed with flux through the flux nozzle 120, and preheated by the preheater 130, and then waved over the solder tank 140. weld. During wave soldering, a liquid solder having a specific shape is formed on the liquid surface of the solder bath 140 by means of a pump by means of a pump. The printed circuit board 170 passes through the crest of the solder wave at a specific angle and a specific immersion depth, thereby realizing the process of solder joint welding.

However, during the wave soldering process, the solder surface may be oxidized to cause the accumulation of solder slag and block the pump, or the consumption during soldering may cause the solder wave height to decrease. The highly unstable solder wave will affect the welding quality, causing problems such as solder point sharpening, air bubbles, pin holes, solder bridge or short circuit. In order to maintain soldering quality, the liquid level of the solder wave needs to be detected quickly and adjusted in real time.

The current detection methods of liquid level are mainly the following three types:

First, use human eyes to visually observe the high temperature resistant glass measuring plate to observe the liquid level. However, this method relies too much on manual judgment, and at the same time, it is necessary to arrange manpower to pay attention to the liquid level change at any time in a high temperature environment.

2. Use the tin wave height measurement special contact sensor sold on the market. The contact type sensor needs to contact the liquid surface, so it is difficult to select an appropriate mounting position. In addition, the measurement range is only the contacted part, and it is impossible to measure the change of the liquid surface height at different positions on the solder wave surface.

3. Use the non-contact sensor special for tin wave height measurement on the market, usually laser and ultrasonic measurement, which need to be installed above the liquid surface. As shown in FIG. 1, a non-contact sensor 180 is installed above the solder tank 140, and emits radar light or ultrasonic waves from the top to the bottom to reach the stable liquid surface to calculate the reflection time difference. For a dynamically fluctuating liquid surface, it cannot provide accurate and accurate Stable level measurement results. Another conventional non-contact sensor is measured by Eddy current, and its measurement distance needs to be within 15mm, so the installation position is limited. And the measurement position is limited to the height change of the liquid surface directly below the sensor, and the height change of the entire liquid surface cannot be obtained.

In view of this, the inventor of the present invention has improved a non-contact method for measuring the height of a liquid surface, and hopes to solve the above-mentioned problems of various conventional methods.

It is an object of the present invention to provide a method for detecting the height of a liquid surface, which has a sensing distance greater than that of a conventional non-contact sensor, and can enable a monitoring person to get away from a high-temperature solder bath while seeing the actual situation on the scene.

Another object of the present invention is to provide a method for detecting a liquid surface height, which can adapt to a dynamically changing liquid surface and accurately measure the liquid surface height of a solder wave.

In order to achieve the above object, the present invention provides a liquid level detection method, which is suitable for detecting a liquid level height of a solder wave rising upward from the inside of a solder tank. The method includes the following steps: first capturing an image to be measured from the side of the solder bath with an image capturing device; selecting an area of interest from the image to be measured; using an image processing unit to obtain brightness data of one of the areas of interest; and An ambient light color; a complex measurement process is provided by an arithmetic unit, and one of the complex measurement processes is selected according to the brightness data and the ambient light color; then, the arithmetic unit executes the selected measurement process to obtain interest A pixel value of the area, which is the number of pixels representing the liquid surface height in the image to be measured; the operation unit provides a conversion coefficient, and uses the conversion coefficient to convert the pixel value into a real height measurement value; and uses A user interface displays actual height measurements.

In one embodiment, the step of providing the conversion coefficient includes: using an image capturing device to capture a correction image from the side of the solder bath, and obtain a correction pixel value and actual liquid level information from the correction image; Then, the arithmetic unit obtains a ratio conversion information of the two based on the corrected pixel value and the actual liquid surface height information, and defines the ratio conversion information as a conversion coefficient.

In one embodiment, the ambient light color has a chromaticity coordinate. The chromaticity coordinate includes a red value, a green value, and a blue value. The sum of the three is a total chromaticity value, and the complex measurement process includes a first A measurement process and a second measurement process. You can use the following judgment steps to decide which measurement process to use. First, provide a brightness reference value and a chromaticity ratio value; then use an arithmetic unit to determine whether the brightness data is less than the brightness reference value; if the brightness data is less than the brightness reference value, provide a light source to illuminate the liquid level of the solder wave, and then An image capture device is used to capture a light image to replace the test image; if the brightness data is greater than the brightness reference value, then determine whether the ratio of the red value to the total chroma value is greater than the chroma ratio; and if the red value accounts for the total If the proportion of the chromaticity value is greater than the proportion of the chromaticity value, the first measurement process is performed, otherwise the second measurement process is performed.

In an embodiment, the first measurement process includes: sequentially performing a frame difference operation, a binarization process, and a morphological process on the region of interest; then, identifying an edge position of the solder wave; and according to the edge Position to obtain the aforementioned pixel value.

In an embodiment, the second measurement process includes: performing contrast enhancement on the region of interest to obtain a contrast-enhanced image to highlight a reflective portion; sequentially performing a gray-scale conversion process, and a binary value on the contrast-enhanced image. Processing, an erosion and expansion process, and an edge detection process; one of the reflective pixel reference values is obtained from the contrast-enhanced image. In addition, contrast-fading is performed on the area of interest to obtain a contrast-fade image to highlight an unreflected portion; the contrast-fade image is processed to a negative to obtain a negative image of the unreflected portion; and the negative image is sequentially An equalization process, a binarization process, an erosion expansion process, and an edge detection process are performed; and an unreflected pixel reference value of one of the unreflected portions is obtained from the negative image. Finally, the reference value of the reflective pixel and the reference value of the non-reflective pixel are compared to obtain the aforementioned pixel value.

In an embodiment, the method further includes: defining a reference value of the liquid level; and comparing the measured value of the actual height with the reference value of the liquid level to obtain a height difference.

In one embodiment, the arithmetic unit obtains a plurality of pixel reference values, which represent the heights of the liquid surface at a plurality of different positions on the entire liquid surface of the solder wave, and uses a conversion coefficient to convert these pixel reference values into a plurality of horizontal height values to form A liquid level wave curve; displaying the liquid level wave curve in the user interface

The invention utilizes non-contact real-time image analysis, which can measure the horizontal height values of multiple points on the entire liquid surface during detection; and provides a mechanism for selecting the measurement process according to the height and ambient color light, so it can adapt to different The dynamic change of liquid level under brightness can also obtain an accurate measurement result under the interference of colored light.

The foregoing and other technical contents, features, and effects of the present invention will be clearly presented in the following detailed description of a preferred embodiment with reference to the accompanying drawings. The directional terms mentioned in the following embodiments, such as: up, down, left, right, front or rear, etc., are merely directions for referring to the accompanying drawings. Therefore, these directional terms are only used to illustrate and not to limit the present invention.

FIG. 2 is a schematic diagram of a liquid level detection system according to an embodiment of the present invention. The liquid level detection system 200 includes an image capturing device 210, an image processing unit 220, an operation unit 230, a user interface 240, and a light source 250. The image capturing device 210 is, for example, a camera, which is suitable for performing image visual processing with the image processing unit 220, and its mounting position is located on the side of the solder tank 140 to obtain an image to be tested. The image processing unit 220 is configured to perform image format conversion, filtering, contrast enhancement or fading, binarization, erosion and expansion, and edge processing on the image to be measured.

The computing unit 230 includes a brightness determination module 232, a chromaticity determination module 234, a pixel interpretation module 236, and a pixel conversion module 238, for performing a determination process and a plurality of different measurement processes. The brightness judging module 232 is configured to judge the brightness of the image to be measured and control the light source 250 accordingly. The chromaticity determination module 234 determines the chromaticity of the image to be measured, and accordingly selects an appropriate measurement process from a variety of measurement processes. The pixel interpretation module 236 is used to determine a pixel value Pm representing the height of the liquid surface in the image to be measured, which is the number of pixels representing the height of the liquid surface in the image to be measured. For example: placing the image to be measured in a pixel In the drawing coordinates as a coordinate unit, at this time, each point on the liquid surface can be represented by a horizontal coordinate and a vertical coordinate, where the number of pixels corresponding to the vertical coordinate is the height of the liquid level representing the point. . The pixel conversion module 238 is used to convert the pixel value Pm representing the liquid surface height in the image to be measured into the liquid surface height of the wave welding site, which is hereinafter referred to as the "real height measurement value", and its unit can be commonly used metric or imperial Unit of length.

In this embodiment, all functions can be established on a user interface 240, for example, as shown in FIG. 2A, a graphical user interface 240 (Graphical User Interface, GUI interface) written by Matlab. The GUI interface 240 can read the information of the image capture device 210 and adjust the function of the image capture device 210, and has the function of reading user settings and storing data, and the GUI interface 240 can display the actual height of the current liquid surface Measured value and current image to be measured. The left half of the GUI interface 240 in FIG. 2A is a setting menu, which includes a plurality of parameter setting fields, and the right half has three display areas 241, 242, and 243. The display area 241 is used to display the current image to be measured; the display area 242 is used to display the liquid level fluctuation curve formed by the horizontal height of each point on the entire liquid surface of the solder wave in the image to be measured; the display area 243 is used to display the solder wave liquid surface Multiple different levels.

The liquid level detection method of the present invention can be performed by using the liquid level detection system 200 described above, including: a screen correction as shown in FIG. 3 and a liquid level reference value HR setting process, a judgment process as shown in FIG. 4, and The liquid level height measurement processes A and B shown in FIGS. 5 to 6. The method of the present invention is suitable for detecting a liquid surface height of a solder wave rising upward from the inside of a solder tank 140, especially when the liquid surface of the solder wave exhibits a dynamic change.

FIG. 3 is a schematic diagram of a screen correction and a liquid level reference value setting method for a liquid level detection method according to an embodiment of the present invention. Before detecting the liquid level, perform a screen correction (Start 1). An image capturing device 210 is used to capture a correction image from the side of the solder bath 140, and the correction image shows a side view of a tin wave rising upward from the solder bath 140 (S300). The image capturing device 210 is an image sensor, such as a charge coupled device (CCD) or a complementary metal oxide semiconductor sensor (CMOS sensor). The image capturing device 210 may have an automatic light fill function, and still maintain a certain clear picture in a dark environment. Before capturing the image for calibration, a standard with a known height, such as a graduated glass plate, needs to be placed on the level of the tin wave surface. The image capture device 210 is then used to capture the calibration image. By calibrating the glass plate, an actual level information Ha of the tin wave can be seen from the image used for the calibration. Next, a function mode is selected (S310), for example, a correction mode M1 or a reference value setting mode M2.

In the correction mode M1, an image capturing device 210 is used to select an image of a region of interest (ROI) from the image for correction, hereinafter referred to as an ROI image. In the calibration mode M1, the measurement distance must be determined, the camera position must be fixed, and then a standard object of known height, such as the aforementioned scale glass plate, is placed at the measurement target to obtain the actual liquid level information Ha. The arithmetic unit 230 uses its pixel interpretation module 236 to obtain a corrected pixel value Pc from the ROI image (S320). The pixel conversion module 238 then obtains the ratio conversion information of the two based on the ratio of the corrected pixel value Pc and the actual liquid level information Ha, and defines the ratio conversion information as a conversion coefficient CF (S322). After calibration, the previously placed standards can be removed.

In the reference value setting mode M2, the correction image is compared with a preset conveying track map (S330) to obtain information of the conveying track 110, such as the height and position of the conveying track 110, or the conveying track 110 The distance from the solder bath 140 is equal. Based on the information of the conveying track 110, a liquid level reference value HR is established (S332). This liquid level reference value HR can be regarded as an ideal liquid level height. The purpose of setting the liquid level reference value HR is to let the user know the difference between the current liquid level and the ideal liquid level.

In one embodiment, a scale-invariant feature transform (SIFT) algorithm can be used to compare the correction image with the transport track map. The SIFT algorithm is a machine vision algorithm used to detect and describe local features in an image. It looks for extreme points in the spatial scale and extracts its position, scale, and rotation invariant numbers. Tolerance for light, noise, and slight viewing angle changes is also quite high. Based on these characteristics, the SIFT algorithm is easy to identify objects with few misidentifications.

After leaving the correction mode M1 or the reference value setting mode M2, it may continue to determine whether to select another function mode (S340). If no other function mode is selected, it is judged whether to start liquid level detection (S350). If yes, first enter a judgment process (Start 2).

FIG. 4 is a schematic flowchart of a method for detecting a liquid surface height according to an embodiment of the present invention. This judging process is used to judge the brightness and the ambient light color of the image to be tested, and then one of them is selected from the complex measurement process for subsequent processing. When the judgment is started (Start 2), an image pre-processing is performed on the image to be tested, and the image pre-processing includes steps S400 to S420:

First, an image capture device 210 is used to capture an image to be tested from the side of the solder tank 140 (S400); then a ROI image is selected from the image to be tested, and the ROI image includes the tin wave side rising from the solder tank 140 View; use the image processing unit 220 to obtain brightness data and an ambient light color of the ROI image; and then judge the brightness and ambient light color of the ROI image.

Since the original ROI image is an RGB format image, it is first converted into a YUV format image (S420), where Y, U, and V are the brightness (Y, Luma), chrominance (U, Chrominance), and density, respectively. (V, Chroma). Then, a first judgment formula is performed on the YUV format image (S430). In a YUV format image, if the Y value is too low, the brightness of the image is very low. If the average Y value is lower than a brightness reference value, for example, the brightness reference value can be set to 35, the brightness judgment module 232 controls the light source 250 to illuminate the liquid surface of the solder wave, and then captures a light with the image capturing device 210 The image replaces the previous image under test (S440). If the Y value is greater than or equal to the brightness reference value on average, it directly jumps to the second judgment formula (S450).

The second judgment formula is to judge the ambient light color of the original ROI image in the RGB format (S450). The light color of a pixel can be represented by a chromaticity coordinate. The chromaticity coordinate includes a red value R, a green value G, and a blue value B, which represent the brightness of red light, green light, and blue light, respectively. The sum of the red value B, the green value G, and the blue value B is called a total chromaticity value. In the original ROI image, if the tones are averaged, ideally, the three values of R, G, and B should each account for 33%. If one of the R, G, and B values is higher, it means that the tone is heavier in the image. In the second judgment formula, if the ratio of the sum of the R values of all pixels in the original ROI image to the sum of the three chromaticity values of R, G, and B of all pixels is greater than a chromaticity ratio value, for example, the chromaticity ratio value is set to 35% That is, if it is determined that there is a red light background, the measurement process A is adopted; if not, the measurement process B is performed.

The above-mentioned two measurement processes of A and B can be built in the arithmetic unit 230 in advance. The arithmetic unit 230 executes the selected measurement process. The pixel interpretation module 236 obtains a pixel value Pm of the ROI image. The pixel value Pm represents the liquid level in the image to be measured. Then the pixel conversion module 238 uses the conversion coefficient CF to convert The pixel value Pm is converted into a real height measurement value Hm at the wave welding site; the real height measurement value Hm is then displayed through the user interface 240.

FIG. 5 is a schematic diagram of a measurement process A of a liquid level detection method according to this embodiment. The measurement process A sequentially performs a filtering process (S500), a frame difference operation (S511 ~ S513 and S521 ~ S522), a binarization process (S531 ~ S532), an AND operation (S540), a Morphological processing (S550) and a marginalization processing (S560).

First, a filtering process is performed on the ROI image (S500) to obtain a filtered image. The filtering process is used to eliminate noise in the ROI image and make the ROI image smoother.

Then, a three-frame-difference method is performed on the filtered image. First take out the Nth, N-1, and N-2 frames (S511 to S513), and then subtract the Nth frame image from the N-1th frame image (S521) to obtain a difference image Diff1; the N-1th The frame image is subtracted from the N-2 frame image (S522) to obtain a differential image Diff2. This operation can find the trajectory of an object by subtracting the number of each frame.

Next, the difference image Diff1 and the difference image Diff2 are binarized to obtain a Diff1 binarized image (S531) and a Diff2 binarized image (S532). Among them, binarization can convert a grayscale image into a binary image. A pixel gray level greater than a certain threshold gray level value is set to a gray level maximum value, and a pixel gray level less than this value is set to a gray level minimum value, thereby achieving binarization. In this embodiment, the first color is black, and the second color is white. Since only the black-and-white image is left in the binarized image, the amount of image calculation can be greatly reduced, so as to increase the efficiency of the image processing unit 220 when the operation is performed and save the waste of memory space.

Then, an AND operation is performed on the Diff1 binary image and the Diff2 binary image to obtain an AND image (S540). Because after binarization, the image has only 1 and 0. In the AND operation, the two are 1, and the output is 1. The remaining conditions are 0. This can be used to obtain the overlapping part of the two frames. Reduce misjudgments.

Morphological processing is performed on the AND image to obtain a morphological image (S550). If the morphological treatment is applied properly, the target can be made more obvious.

Next, edge processing is performed on the morphological image to obtain an edged image (S560). The purpose of edge detection processing is to find the boundary between the tin wave and the background. The detection of edges is mostly determined by the difference in the grayscale values of adjacent pixels in the image. If the grayscale value varies greatly, it is the edge position. And vice versa. However, in many cases, the edge position will not be located at just one pixel, but will be composed of several pixels, and the true edge position is among these pixels.

After obtaining this marginalized image, the entire ROI image is regarded as a matrix, and the characteristics of binarization non-zero or 1 are used to obtain a plurality of liquid surface heights representing a plurality of different positions of the tin wave liquid surface in the ROI image. Pixel reference value Pn (S570). In a preferred embodiment, the extreme values of these pixel reference values Pn can be removed by curve fitting to avoid multi-point offset (S580) to obtain the required pixel value Pm and reduce the error value. In this embodiment, curve fitting refers to using a continuous curve to roughly describe or compare the functional relationship between the coordinates represented by the discrete point groups on the plane.

This pixel value Pm is then subjected to a pixel value conversion process according to the conversion coefficient CF obtained in the previous calibration process, and the measured pixel value Pm can be converted into a real height measurement value Hm, the unit of which is mm, and this real height The measured value Hm represents the current tin wave level (S590). Finally, the actual height measurement value Hm is subtracted from the predetermined liquid level reference value HR to obtain the height difference between the current liquid level and the ideal liquid level. It is worth mentioning that the method of the present invention can convert the pixel reference values Pn obtained in the previous step S570 into pixel values, and then obtain the height values of multiple different points on the entire surface of the tin wave. The fluctuation of the overall liquid level of the wave is shown in the GUI interface 240 in FIG. 2A, which is a function not available in the conventional technology.

FIG. 6 is a schematic diagram of a measurement process B of the liquid level detection method according to this embodiment. Due to the problem of light and shadow inside the solder bath 140, the tin wave may be partially reflected, so that the tin wave has a different color. In this embodiment, two different processes are performed on the reflective part and the non-reflective part of the ROI image, and finally a formula is used to calculate the measured tin wave height.

The left half (S610 to S615) of Fig. 6 is to deal with the tin wave in the reflective part, which tin light belongs to the light color. First, the ROI image is contrast-enhanced to obtain a contrast-enhanced image to highlight a reflective portion (S610). The step of contrast enhancement is to increase the brightness of a specific pixel. The purpose is to highlight the difference between the tin wave in the reflective part and the background, so as to reduce the degree of error caused by subsequent image processing.

Secondly, a gray-scale conversion process is performed on the contrast-enhanced image of the reflective portion to obtain a reflective gray-scale image (S611). This is because color images have quite a lot of different information. In order to effectively reduce the amount of information in the image and the complexity represented by the graphics, grayscale conversion processing is usually performed.

Then, the reflective grayscale image is binarized to obtain a reflective binary image (S612). Then, the reflective binarized image is subjected to an erosion expansion process (S613). Erosion and swelling is a way to eliminate noise, and you need to set a mask first. The result of erosion calculation will make the image look shrunk, and unnecessary elements can be removed by using appropriate structural elements. The result of the expansion calculation will make the image look larger, and the gap can be filled with appropriate structural elements.

Then, an edge detection process is performed on the image after the erosion and expansion process to obtain a reflective edge enhanced image (S614). The entire reflective edge-enhanced image is regarded as a matrix, and a plurality of reflective pixel reference values PR1 (S615) are obtained by binarizing non-zero or one (S615), which represent multiple different positions on the tin wave liquid surface of the reflective portion Liquid level. In a preferred embodiment, the aforementioned curve fitting can be used to avoid multi-point offsets to reduce the error value.

The right half (S620 to S626) of Fig. 6 belongs to the non-reflective part. The tin wave in this part is black.

First, contrast-fading the ROI image to obtain a contrast-fading image to highlight an unreflected portion (S620). This method is because the background of the image is too dim, so that the subsequent binarization can effectively distinguish the tin wave from its unreflected part and its background. Then, the contrast-faded image is subjected to a negative film processing, and the colors of the black and white pixels are interchanged. In this embodiment, the original white pixels are converted into black pixels to obtain a negative image of the unreflected portion (S621).

Then, the negative image of the unreflected part is subjected to histogram equalization to obtain an unreflected equalized image (S622). The probability of the grayscale values of the image is not equal, and the histogram equalization is to transform the grayscale value of the image, so that the probability of the grayscale value after the change is the same, the main purpose is to make the image Detailed information can be more prominent.

Then, a grayscale conversion process is performed on the unreflected equalized image to obtain an unreflected grayscale image. Then a binarization process is performed on the unreflected grayscale image to obtain an unreflected binarized image (S623).

An erosion and expansion process is performed on the unreflected binary image (S624) to obtain an unreflected erosion and expansion image. Different from the process of erosion and expansion of the reflective part, the background inside the solder bath 140 will be fused with the characteristics of tin waves due to the reflection, which makes the identification of tin waves difficult. Therefore, a mask must be provided for the processing of the reflective part. To adjust the vertical segments. The subsequent calculation of erosion and expansion is the same as that of the reflective part.

Subsequently, an edge detection process is performed on the unreflected erosion and expansion processed image to obtain an unreflected edge enhanced image (S625), and 0 and 1 are identified to find the tin wave liquid level height (S626). The entire unreflected edge-enhanced image is regarded as a matrix, and the non-reflective pixel reference value PR2 is obtained by binarizing non-zero or 1, which represents a plurality of different positions on the tin wave surface of the unreflected portion. Liquid level. In a preferred embodiment, the aforementioned curve fitting can be used to avoid multi-point offsets to reduce the error value.

Then, the plurality of reflective pixel reference values PR1 and the plurality of non-reflective pixel reference values PR2 obtained by the two processing methods are compared (S630) to obtain a plurality of pixel reference values Pn. The required pixel value Pm is obtained by eliminating the pixel reference value Pn that is too low and too high (S640), and then the pixel value is converted according to the conversion coefficient CF obtained in the previous calibration process (S650), The pixel value Pm is converted into the actual height measurement value Hm, and the unit is mm. This actual height measurement Hm is the current tin wave level. Finally, the actual height measurement value Hm is subtracted from the liquid level reference value HR to obtain the height difference Hd between the two, which represents the difference between the current liquid level and the preset ideal liquid level (S660).

To sum up, the present invention provides a method for detecting the height of a liquid surface by using an image. Through an imaging device with a lens, the dynamic / static liquid surface is photographed, and an image frame is captured in a side manner. Perform image processing in a specific range to reduce unnecessary calculations. Use image processing technology to detect the edge of the liquid surface and obtain the average value of the fluctuation of the liquid surface. The pixel value representing the height of the liquid surface can be accurately measured and removed. The maximum or minimum pixel value is then converted into the liquid surface height of the wave welding site through linear regression of equal proportions.

In addition, the invention performs real-time image processing on the tin wave height captured by the camera, so that the user can see the tin wave height change status of the current wave welding site on the monitor, so that the factory can quickly grasp the correct information in management and maintenance , Which in turn improves the soldering quality of printed circuit boards. This product does not need to restrict specific photographic lenses, and can be used with general consumer photography equipment. Especially in factories with high temperature environment, the advantages of this method can be more prominent.

However, the above are only the preferred embodiments of the present invention. When the scope of implementation of the present invention cannot be limited by this, that is, the simple equivalent changes and modifications made according to the scope of the patent application and the description of the invention, All are still within the scope of the invention patent. In addition, any embodiment of the present invention or the scope of patent application does not need to achieve all the purposes or advantages or features disclosed by the invention. In addition, the abstract and the title are only used to assist the search of patent documents, and are not intended to limit the scope of rights of the present invention.

100‧‧‧The conventional wave welding system

110‧‧‧Conveyor track

120‧‧‧Flux Nozzle

130‧‧‧preheater

140‧‧‧solder bath

150‧‧‧fan

160‧‧‧ Central controller

170‧‧‧printed circuit board

172‧‧‧Dual In-line Package Components

180‧‧‧ Non-contact sensor

200‧‧‧Liquid level detection system

210‧‧‧Image capture device

220‧‧‧Image Processing Unit

230‧‧‧ Computing Unit

232‧‧‧Brightness judgment module

234‧‧‧Chroma Judgment Module

236‧‧‧Pixel Interpretation Module

238‧‧‧pixel conversion module

240‧‧‧user interface (GUI interface)

241, 242, 243‧‧‧ display area

250‧‧‧ light source

S300 ~ S350‧‧‧Screen calibration and liquid level reference value setting process

M1‧‧‧ Calibration mode

M2‧‧‧reference value setting mode

Ha‧‧‧ Actual Level Information

Pc‧‧‧corrected pixel value

CF‧‧‧ Conversion Factor

HR‧‧‧Basic level height reference value

S400 ~ S450‧‧‧Judging process

(S500 ~ S590) ‧‧‧Measurement process A

(S610 ~ S660) ‧‧‧Measurement Process B

ROI‧‧‧ Area of Interest

R‧‧‧ red value

G‧‧‧Green value

B‧‧‧ blue value

Y‧‧‧Brightness

U‧‧‧ Chroma

V‧‧‧ concentration

Diff1, Diff2‧‧‧ Differential images

PR1‧‧‧reflective pixel reference value

PR2‧‧‧unreflected pixel reference value

Pn‧‧‧pixel reference value

Pm‧‧‧pixel value

Hd‧‧‧height difference

Hm‧‧‧Real height measurement

FIG. 1 is a schematic diagram of a conventional wave welding system.

FIG. 2 is a schematic diagram of a liquid level detection system according to an embodiment of the present invention.

FIG. 2A is a schematic diagram of a user interface according to an embodiment of the present invention.

FIG. 3 is a schematic diagram of a screen correction and a liquid level reference setting process according to an embodiment of the present invention.

FIG. 4 is a schematic flowchart of a method for detecting a liquid surface height according to an embodiment of the present invention.

FIG. 5 is a schematic diagram of a measurement process A according to an embodiment of the present invention.

FIG. 6 is a schematic diagram of a measurement process B according to an embodiment of the present invention.

no

no

Claims (10)

A liquid surface height detection method is suitable for detecting a liquid surface height of a solder wave rising upwardly from the inside of a solder tank. The method includes: using an image capturing device to capture an image from a side of the solder tank to be Measuring an image; selecting an area of interest from the image to be measured; using an image processing unit to obtain brightness data and an ambient light color of the area of interest; providing a quantitative measurement process with an arithmetic unit, and according to the brightness data And the ambient light color, select one of the plurality of measurement processes; the arithmetic unit executes the selected measurement process to obtain a pixel value of the region of interest, which represents the liquid in the image to be measured The number of pixels of the surface height; the operation unit provides a conversion coefficient, and uses the conversion coefficient to convert the pixel value into a real height measurement value; and uses a user interface to display the real height measurement value. The liquid level detection method according to item 1 of the scope of the patent application, wherein the step of providing the conversion coefficient includes: using the image capturing device to capture an image for calibration from the side of the solder bath; Obtain a corrected pixel value and an actual liquid surface height information from the image; and the arithmetic unit obtains a ratio conversion information of the two based on the corrected pixel value and the actual liquid surface height information, and converts the ratio conversion information Defined as the conversion factor. The method for detecting the height of a liquid surface according to item 1 of the scope of patent application, wherein the ambient light color has a chromaticity coordinate, and the chromaticity coordinate includes a red value, a green value, and a blue value, wherein the red value, The sum of the green value and the blue value is a total chromaticity value, and the complex measurement process includes a first measurement process and a second measurement process, and one of the plurality of measurement processes is selected from the plurality of measurement processes. The steps include: providing a brightness reference value and a chromaticity ratio value; using the arithmetic unit to determine whether the brightness data is smaller than the brightness reference value; and if the brightness data is greater than the brightness reference value, determining whether the red value accounts for the total chromaticity Whether the value ratio is greater than the chroma ratio value; and if the ratio of the red value to the total chroma value is greater than the chroma ratio value, the first measurement process is performed, otherwise the second measurement process is performed to Get the pixel value. The liquid surface height detection method according to item 3 of the scope of patent application, wherein if the brightness data is less than the brightness reference value, a light source is provided to illuminate the liquid surface of the solder wave, and then the image capture device is used to capture Take a light image to replace the test image. The liquid level detection method according to item 3 of the patent application scope, wherein the region of interest includes a side view of the solder wave, and the first measurement process includes: identifying an edge position of the solder wave; and The edge position obtains the pixel value. The liquid level detection method according to item 5 of the scope of patent application, wherein the first measurement process includes: before identifying the edge position, sequentially performing a frame difference operation and a binarization on the region of interest Treatment and a morphological treatment. The liquid level detection method according to item 3 of the scope of patent application, wherein the second measurement process includes: contrast-enhancing the region of interest to obtain a contrast-enhanced image to highlight a reflective portion; A reference value for one reflective pixel of the reflective portion is obtained in the contrast-enhanced image; a contrast-fading image is obtained for the region of interest to highlight a non-reflective portion; A negative image of the unreflected portion; obtaining an unreflected pixel reference value of the unreflected portion from the negative image; and comparing the reflected pixel reference value and the unreflected pixel reference value to obtain the pixel value. The liquid level detection method according to item 7 of the scope of the patent application, wherein the second measurement process includes: sequentially performing a gray-scale conversion process on the contrast-enhanced image, a binarization process, and an erosion expansion Processing and an edge detection process; and sequentially performing an equalization process, a binarization process, an erosion and expansion process, and an edge detection process on the negative image. The method for detecting the liquid level according to item 1 of the patent application scope further includes: defining a reference value for the liquid level; and comparing the measured value of the actual height with the reference value for the liquid level to obtain a height difference. The liquid surface height detection method according to item 1 of the patent application scope, wherein the region of interest includes the entire liquid surface of the solder wave, and the method further includes: the arithmetic unit obtains a reference value of a plurality of pixels from the region of interest Wherein the reference value of the plurality of pixels represents the height of the liquid surface at a plurality of different positions on the entire liquid surface of the solder wave; using the conversion coefficient to convert the reference values of the pixels to a plurality of horizontal height values to form a liquid level wave curve; and The liquid level wave curve is displayed in the user interface.