Layering defect detection method and system based on integrated circuit packaging

Abstract

The invention discloses a layering defect detection method and system based on integrated circuit packaging, and relates to the technical field of integrated circuit packaging detection.The main scheme is that ultrasonic frequency is calculated by accurately obtaining material and defect size data, layering defects can be effectively detected, and the accuracy of a detection result is improved; the method comprises the following steps: transmitting ultrasonic waves conforming to an ultrasonic frequency interval to an integrated circuit package through an ultrasonic transmitting device, receiving an ultrasonic signal, constructing an internal structure image of the integrated circuit package, and analyzing; according to the method, whether gray values of adjacent pixel points of an internal structure image are abnormal or not is judged, whether colors of the adjacent pixel points are continuously abnormal or not is judged, gray value abnormal data and color abnormal data are generated respectively, and whether the integrated circuit package has a layering defect or not is judged according to the abnormal data. And a detailed report is generated, so that quality monitoring and improvement in the production process are facilitated, and the overall quality of the product is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of integrated circuit package detection, and specifically to a method and system for detecting delamination defects based on integrated circuit packages. Background Art

[0002] With the continuous development of integrated circuit technology, the complexity and density of integrated circuit packages have been increasing day by day. Package delamination defects have become one of the important factors affecting the performance and reliability of integrated circuits. Traditional detection methods, such as visual inspection and electrical testing, often have difficulty in accurately detecting tiny delamination defects.

[0003] Traditional detection methods cannot calculate the accurate ultrasonic frequency based on material and defect size data; therefore, it is difficult to effectively detect delamination defects, reducing the accuracy of detection results; the analysis of internal structure images is slow, and it is necessary to damage the package to detect potential defects, increasing the detection cost and time cost and reducing production efficiency; it cannot classify the quality level of packages, nor can it generate a detailed report, which is not conducive to quality monitoring and improvement in the production process, making it difficult to improve the overall quality of products. Summary of the Invention

[0004] (1) Technical Problems to be Solved

[0005] Aiming at the deficiencies of the prior art, the present invention provides a method and system for detecting delamination defects based on integrated circuit packages, which can detect tiny delamination defects. By analyzing the gray value and color continuity of adjacent pixel points in the internal structure image, very subtle anomalies can be captured, improving the detection ability for tiny delamination defects and solving the problem that traditional detection methods, such as visual inspection and electrical testing, often have difficulty in accurately detecting tiny delamination defects.

[0006] (2) Technical Solutions

[0007] To achieve the above objectives, the present invention is realized through the following technical solutions: A method for detecting delamination defects based on integrated circuit packages, including:

[0008] Obtaining the material data and defect size data of an integrated circuit package; calculating the ultrasonic frequency range matching the material data and defect size data;

[0009] Emitting ultrasonic waves conforming to the ultrasonic frequency range to the integrated circuit package through an ultrasonic wave emitting device; receiving the ultrasonic wave signal and constructing an internal structure image of the integrated circuit package;

[0010] Analyze the internal structure image to determine whether there are abnormalities in the gray values of adjacent pixels in the internal structure image and whether there are continuity abnormalities in the colors of adjacent pixels, and generate gray value abnormal data and color abnormal data respectively. Determine whether there are delamination defects in the integrated circuit package according to the gray value abnormal data and color abnormal data.

[0011] According to the gray value abnormal data and color abnormal data of multiple integrated circuit packages, classify the quality level of each integrated circuit package and generate a production quality level report.

[0012] In the preferred solution of the above delamination defect detection method based on integrated circuits: The method for calculating the ultrasonic frequency range is as follows:

[0013] Select the wave speed range data that matches it according to the material data ;

[0014] Calculate the wavelength data according to the defect size data , specifically:

[0015] The defect size data includes at least the minimum delamination defect size , according to the minimum delamination defect size Calculate the wavelength data , and the formula is as follows:

[0016]

[0017] According to the wave speed range data and the wavelength data Calculate the ultrasonic frequency range , and the formula is as follows:

[0018] .

[0019] In the preferred solution of the above delamination defect detection method for circuit packages: The method for constructing the internal structure image is as follows:

[0020] Obtain the ultrasonic intensities reflected at different measurement positions , and calculate the gray values at different measurement positions through the ultrasonic intensities , and the formula is as follows: , where:

[0021]

[0022] Among them: and are constants, and by mapping the ultrasonic intensity to the gray value ;

[0023] When constructing an image, different measurement positions are mapped to the pixel positions of the image, and the th measurement position corresponds to the th pixel in the image; the calculated gray value is assigned to this pixel.

[0024] In the preferred solution of the above-mentioned hierarchical defect detection method based on integrated circuit packaging: The method for calculating the intensity of the received ultrasonic signal is as follows:

[0025] Obtain the ultrasonic transmission power and the distance of ultrasonic propagation , and calculate the intensity of the received ultrasonic signal through the ultrasonic transmission power and the distance of ultrasonic propagation . The formula is as follows:

[0026] .

[0027] In the preferred solution of the above-mentioned hierarchical defect detection method based on integrated circuit packaging: The method for calculating the ultrasonic transmission power is as follows:

[0028] Obtain the medium density , the sound speed , the vibration area , the angular frequency and the amplitude , and calculate the ultrasonic transmission power through the medium density , the sound speed , the vibration area and the amplitude . The formula is as follows:

[0029] .

[0030] In the preferred solution of the above-mentioned hierarchical defect detection method based on integrated circuit packaging: The abnormal gray value data includes: the average gray value difference and the proportion of abnormal gray value pixels ; among them, the calculation method of the average gray value difference is as follows:

[0031] Obtain the gray values of several pixel points as and the gray values of the adjacent pixel points as , and through the pixel point The grayscale value is and its adjacent pixel points The grayscale value is Calculate the difference in grayscale values of adjacent pixel points , and the formula is as follows:

[0032] ;

[0033] Based on the difference in grayscale values of adjacent pixel points , calculate the average difference in grayscale values , and the formula is as follows:

[0034] ;

[0035] Where: is the th difference in grayscale values of adjacent pixel points, and the value range of is 1, 2, 3... is the number of pairs of adjacent pixel points selected , and the value is a positive integer;

[0036] The proportion of pixel points with abnormal grayscale values is calculated as follows:

[0037] Set the threshold for determining abnormal grayscale changes ; The method for calculating the threshold for determining abnormal grayscale changes is as follows:

[0038] ;

[0039] Where: is the coefficient of the average difference in grayscale values , and when , it is determined that the grayscale change of this pair of pixel points is abnormal;

[0040] Obtain the number of pairs of pixel points with abnormal grayscale values , and calculate the proportion of pixel points with abnormal grayscale values based on the number of pairs of pixel points with abnormal grayscale values using the following formula:

[0041] ;

[0042] Where: is the total number of pairs of adjacent pixel points.

[0043] In the preferred solution of the above method for detecting hierarchical defects in integrated circuit packaging: The color anomaly data includes: the average hue difference , the average saturation difference , the average value of brightness difference and the proportion of color abnormal pixel points ; among them, the average value of hue difference , the average value of saturation difference and the average value of brightness difference are calculated as follows:

[0044] The method for judging whether there is a continuity abnormality in the colors of adjacent pixel points is as follows:

[0045] Let be the red, green, and blue color component values of a certain pixel point in the RGB color space respectively, and their value ranges are all [0, 255]; calculate ;

[0046] represents the maximum value among the red, green, and blue color components of the pixel point, represents the minimum value among the red, green, and blue color components of the pixel point,

[0047] Then calculate the hue :

[0048] When , ;

[0049] When and , ;

[0050] When and , ;

[0051] When , ;

[0052] When , ;

[0053] Calculate the saturation :

[0054] When , ;

[0055] When , ;

[0056] Calculate the brightness ;

[0057] Let a certain pixel point in the HSV color space be , , The values of its adjacent pixel points in the HSV color space are 、 、 ;

[0058] The formula for calculating the hue difference is as follows:

[0059] ;

[0060] The formula for calculating the saturation difference is as follows:

[0061] ;

[0062] The formula for calculating the value difference is as follows:

[0063] ;

[0064] Based on the hue differences 、saturation differences and value differences of several pairs of pixel points, calculate the average hue difference 、average saturation difference and average value difference , and the formulas are as follows:

[0065] ;

[0066] Where: is the number of pairs of adjacent pixel points selected, is the hue difference of the th pair of adjacent pixel points, ranges from 1, 2, 3...d, where d is a positive integer, is the saturation difference of the th pair of adjacent pixel points, is the value difference of the th pair of adjacent pixel points;

[0067] The method for calculating the proportion of color abnormal pixel points is as follows:

[0068] Preset color discontinuity determination, hue threshold 、saturation threshold 、value threshold , when 、 、 , if any one of the three is satisfied, it is determined that this pair of pixel points is color discontinuous in the corresponding attribute. Where: Is the hue difference The weight coefficient ranges from 0.2 to 0.3. is the saturation difference Weight coefficient, value range 0.3~0.4, Is the brightness difference The weight coefficient ranges from 0.3 to 0.4. + + =1;

[0069] Get the number of color abnormal pixels , by the number of color abnormal pixels Calculate the ratio of color abnormal pixels The formula is as follows:

[0070] .

[0071] In the preferred embodiment of the above-mentioned method for detecting delamination defects based on integrated circuit packaging: the method for determining whether the integrated circuit package has delamination defects is:

[0072] The specific method for determining whether the gray value abnormal data is abnormal is:

[0073] The mean gray value difference Gray value change abnormality judgment threshold For comparison, When the average gray value difference is determined abnormal;

[0074] Set the gray value abnormal ratio threshold ;

[0075] The ratio of abnormal grayscale pixels Gray value abnormal ratio threshold Compare, when When the gray value is abnormal, the proportion of pixels is determined abnormal;

[0076] When the mean gray value difference Ratio of abnormal pixels to gray value If one of the items is abnormal, it is determined that the gray value abnormal data is abnormal;

[0077] The specific method for determining whether color abnormality data is abnormal is:

[0078] Average the hue differences With hue threshold , Saturation difference average With saturation threshold , average value of brightness difference Compare with the brightness threshold When 、 、 If one of the three cases is abnormal, it is determined that the color of this pixel pair is discontinuous in the corresponding attribute;

[0079] Set the threshold of the color abnormality ratio ;

[0080] Compare the ratio of color abnormal pixel points with the threshold of the color abnormality ratio When it is determined that the ratio of color abnormal pixel points is abnormal;

[0081] When the average value of hue difference 、the average value of saturation difference 、the average value of brightness difference and the ratio of color abnormal pixel points If one of them is abnormal, it is determined whether the color abnormal data is abnormal;

[0082] When one of the gray value abnormal data and the color abnormal data is abnormal, it is determined that the integrated circuit package has a delamination defect. When the integrated circuit package has a delamination defect, a warning is issued.

[0083] In the preferred scheme of the above method for detecting delamination defects of integrated circuit packages: The method for generating a production quality grade report is as follows:

[0084] Obtain the average gray value difference of each normal product 、the ratio of gray value abnormal pixel points 、the average value of hue difference 、the average value of saturation difference 、the average value of brightness difference and the ratio of color abnormal pixel points , calculate the quality score , and the formula is as follows:

[0085] ;

[0086] Among them: is the weight coefficient of the average gray value difference , and the value range is 0.1~0.2, is the weight coefficient of the ratio of gray value abnormal pixel points , and the value range is 0.2~0.3, is the weight coefficient of the average value of hue difference , and the value range is 0.1~0.2, is the average value of saturation difference is the weight coefficient, with a value range of 0.2 to 0.3, is the average value of lightness difference is the weight coefficient, with a value range of 0.1 to 0.2, is the proportion of color abnormal pixel points is the weight coefficient, with a value range of 0.2 to 0.3, and + + + + + = 1;

[0087] Set the quality score threshold for first-class products as , when it is, it is a first-class product; the quality score threshold for first-class products is , , it is a first-class product; the quality score threshold for qualified products is , when it is, it is a qualified product;

[0088] Count the total number of products of integrated circuit packages , the number of first-class products , the number of first-class products , the number of qualified products and the number of unqualified products ; Calculate the proportion of first-class products , the proportion of first-class products , the proportion of qualified products and the proportion of unqualified products The formulas are as follows:

[0089] ;

[0090] Preset the unqualified product threshold , when the unqualified proportion > the unqualified product threshold issue an alarm;

[0091] Generate a production quality grade report according to the proportion of first-class products , the proportion of first-class products , the proportion of qualified products and the proportion of unqualified products

[0092] The present invention also discloses a hierarchical defect detection system based on integrated circuit packaging, a data acquisition and frequency calculation module: acquiring the material data and defect size data of an integrated circuit package, and calculating the ultrasonic frequency range matching the material data and defect size data; ​

[0093] The ultrasonic emission module emits ultrasonic waves within the ultrasonic frequency range towards the integrated circuit package through an ultrasonic emission device;

[0094] The signal reception and image construction module receives ultrasonic signals and constructs an internal structure image of the integrated circuit package;

[0095] The defect analysis module analyzes the internal structure image, determines whether there are abnormalities in the gray values of adjacent pixel points in the internal structure image and whether there are continuity abnormalities in the colors of adjacent pixel points, generates gray value abnormality data and color abnormality data respectively, and determines whether there are delamination defects in the integrated circuit package based on the gray value abnormality data and the color abnormality data;

[0096] The quality grading and report generation module: Based on the gray value abnormality data and color abnormality data of multiple integrated circuit packages, conducts quality grade classification for each integrated circuit package and generates a production quality grade report.

[0097] (III) Beneficial effects

[0098] The present invention provides the patent name and has the following beneficial effects:

[0099] (1) Improve detection accuracy: By accurately obtaining material and defect size data to calculate the appropriate ultrasonic frequency, it is possible to more effectively detect delamination defects and improve the accuracy of detection results;

[0100] (2) Enhance quality control: It can conduct quality grade classification for packages and generate detailed reports, which helps in quality monitoring and improvement during the production process, thereby enhancing the overall quality of the product;

[0101] (3) Reduce production costs: By quickly analyzing the internal structure image, potential delamination defects can be detected in a timely manner without damaging the package, reducing detection costs and time costs and improving production efficiency;

[0102] (4) Optimize detection efficiency: Each module has a clear division of labor and works collaboratively, reducing redundant steps and time waste during the detection process and improving the overall detection efficiency. Description of the drawings

[0103] Figure 1 It is a step schematic diagram of a delamination defect detection method based on an integrated circuit package; Specific implementation manners

[0104] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0105] See also Figure 1 The present invention provides a delamination defect detection method based on integrated circuit packaging, comprising the following steps:

[0106] Step 1: Obtain material data and defect size data of the integrated circuit package; calculate the matching ultrasonic frequency range according to the material data and defect size data;

[0107] Step 1 includes the following:

[0108] Step 101: The method for calculating the ultrasonic frequency range is as follows:

[0109] Select the wave velocity interval data that matches the material data ;

[0110] Calculate wavelength data based on defect size data , specifically:

[0111] Defect size data includes at least the minimum delamination defect size , according to the minimum delamination defect size Calculating wavelength data , the formula is as follows:

[0112] ;

[0113] Where: Minimum delamination defect size The specific value is determined in combination with actual production and quality control requirements;

[0114] According to the wave speed interval data and wavelength data Calculate the ultrasonic frequency range , the formula is as follows:

[0115] ;

[0116] It should be noted that for commonly used materials for integrated circuit packaging, the wave speed There is a certain range, It is the propagation speed of ultrasound in the material, and different packaging materials have different values; is the wavelength in ultrasonic frequency materials, is the ultrasonic frequency, and the appropriate frequency range needs to be determined here to ;

[0117] According to the detection requirements, the wavelength should match the size of the possible delamination defects. Generally speaking, the wavelength should be less than half of the defect size in order to detect the defects better. For example, if the minimum delamination defect size to be detected is , then ; Assuming the minimum defect size ;

[0118] Wave velocity data needs to be selected according to the material. For example, for plastic encapsulation materials:

[0119] When , , that is ;

[0120] When , ;

[0121] For ceramic encapsulation materials:

[0122] When , that is ;

[0123] When , ;

[0124] Taking all factors into consideration, take , ;

[0125] By accurately obtaining the material and defect size data to calculate the appropriate ultrasonic frequency, the delamination defects can be detected more effectively, the accuracy of the detection results is improved, the delamination defects are detected accurately at an early stage, the unqualified products are prevented from flowing into the subsequent production links, and the material waste and production costs are reduced.

[0126] Step 2: Transmit ultrasonic waves that meet the ultrasonic frequency range to the integrated circuit package through the ultrasonic wave transmitting device; receive the ultrasonic wave signals and construct an internal structure image of the integrated circuit package;

[0127] Step 2 includes the following contents:

[0128] Step 201: Obtain the ultrasonic intensities reflected at different measurement positions , and calculate the gray values at different measurement positions through the ultrasonic intensity , and the formula is as follows:

[0129] ​;

[0130] Wherein: and are constants. By mapping the intensity of ultrasonic waves to the gray value , represents the gray value of the pixel at the coordinate . The value range is [0, 255]. Different reflection intensities can be represented by different grays in the image, reflecting different material or structural characteristics inside the integrated circuit package;

[0131] It should be noted that the gray value obtained through formula calculation can construct an image reflecting the internal structural characteristics of the package, providing a basis for detecting delamination defects; in practical applications, the parameters can be further optimized and adjusted according to the specific device performance and detection requirements.

[0132] Step 202: Obtain the ultrasonic transmission power and the distance that the ultrasonic wave propagates . Calculate the intensity of the received ultrasonic signal through the ultrasonic transmission power and the distance that the ultrasonic wave propagates . The formula is as follows:

[0133] ;

[0134] Wherein: The distance that the ultrasonic wave propagates is measured according to the actual detection setting; the value of the intensity of the received ultrasonic signal depends on the transmission power and the propagation distance, and usually varies within a certain intensity range; the ultrasonic transmission power is calculated according to the capabilities of the transmitting device and the detection requirements, and is usually selected within a specific power range. The distance that the ultrasonic wave propagates is measured according to the actual detection setting.

[0135] Step 203: Obtain the medium density , the sound velocity , the vibration area , the angular frequency and the amplitude . Calculate the ultrasonic transmission power through the medium density , the sound velocity , the vibration area and the amplitude . The formula is as follows:

[0136] ​​

[0137] Among them: medium density and sound velocity depend on the package material, vibration area is determined by the probe, angular frequency is determined by the set frequency, amplitude is controlled by the transmitting device;

[0138] It should be noted that based on the principle of acoustic wave energy propagation, considering the medium density and sound velocity , vibration area , angular frequency and amplitude to calculate the transmission power.

[0139] Step 3: Analyze the internal structure image, determine whether there are abnormalities in the gray values of adjacent pixel points in the internal structure image and whether there are continuity abnormalities in the colors of adjacent pixel points, respectively generate gray value abnormal data and color abnormal data, and determine whether there are delamination defects in the integrated circuit package according to the gray value abnormal data and color abnormal data;

[0140] Step 3 includes the following content:

[0141] Step 301: The gray value abnormal data includes: average gray value difference and the proportion of gray value abnormal pixel points ; Among them, the calculation method of the average gray value difference is as follows:

[0142] Obtain the gray values of several pixel points from the internal structure image as and the gray value of the adjacent pixel point as , through the pixel point with the gray value of and its adjacent pixel point with the gray value of calculate the gray value difference of adjacent pixel points , and the formula is as follows:

[0143] ;

[0144] Among them: the gray value difference of adjacent pixel points is a non-negative value, and its value range is [0, +∞), but in the actual image, due to the certain range of gray values, its difference value will also be within a certain range;

[0145] It should be noted that the degree of change in the gray values of adjacent pixel points is measured by calculating the absolute value of the difference in the gray values of adjacent pixel points. If the difference is large, it indicates that the gray value changes violently, and there may be a delamination defect, resulting in a change in the material or structural properties, thus causing a change in the intensity of the reflected ultrasonic wave, which is then reflected in the change in the gray value.

[0146] Step 302: Calculate the average gray value difference through the gray value difference of adjacent pixel points , and calculate the average gray value difference . The formula is as follows:

[0147] ;

[0148] Where: The average gray value difference is a statistical value calculated based on the selected pixel point pairs, and its value range depends on the value range of the gray value difference of adjacent pixel points ;

[0149] It should be noted that first, the overall gray change of the image is obtained by calculating the average gray value difference, and then a threshold related to the average gray value difference is set. When the gray value difference of a certain pixel point pair exceeds this threshold, it is considered that the gray change in this area is abnormal. This is based on the assumption that in a defect-free area, the gray value changes relatively little and evenly, while in a defective area, the gray value has a large mutation.

[0150] Step 303: Set the threshold for determining abnormal gray change ; The calculation method of the threshold for determining abnormal gray change is as follows: The calculation method is as follows:

[0151] ;

[0152] Where: The value of the threshold for determining abnormal gray change will change with the coefficient of the average gray value difference and the average gray value difference . The value of the coefficient of the average gray value difference should be determined according to the requirements for the sensitivity and specificity of defect detection. If the value is large, the determination of abnormal gray change will be more strict, which may reduce false positives but may also miss some defects; if the value is small, the determination of abnormal gray change will be more lenient, which may detect more suspected defects but may also increase the probability of false positives.

[0153] Step 304: Obtain the number of pixel point pairs with abnormal gray values , and calculate the proportion of pixel points with abnormal gray values through the number of pixel point pairs with abnormal gray values ​ The formula is as follows:

[0154] ;

[0155] Among them: the number of pixel pairs with abnormal gray values is obtained by judging and counting each pixel according to the previously set criterion for judging abnormal gray values; for example, if in a detection area containing 1000 pixels, 200 pixels are judged to have abnormal gray values, then the proportion of pixels with abnormal gray values ; the meaning of the formula is to calculate the percentage of pixels with abnormal gray values in the entire image or detection area among the total number of pixels.

[0156] Step 305: The color anomaly data includes: the average hue difference , the average saturation difference , the average lightness difference and the proportion of pixels with color anomalies ; among them, the calculation methods of the average hue difference , the average saturation difference and the average lightness difference are as follows:

[0157] The method for judging whether there is a continuity anomaly in the colors of adjacent pixels is as follows:

[0158] Let be the red, green, and blue color component values of a certain pixel in the RGB color space respectively, and their value ranges are all [0, 255]; calculate ;

[0159] represents the maximum value among the red, green, and blue color components of the pixel, represents the minimum value among the red, green, and blue color components of the pixel,

[0160] Then calculate the hue :

[0161] When , ;

[0162] When and , ;

[0163] When and , ;

[0164] When , ;

[0165] When it is ;

[0166] Calculate the saturation :

[0167] When it is ;

[0168] When it is ;

[0169] Calculate the lightness ;

[0170] It should be noted that: The RGB color space is a color representation method based on the mixing of the three primary colors of red, green, and blue. However, it does not intuitively reflect the characteristics of colors. The HSV color space divides colors into hue , saturation and lightness in three attributes, which is more in line with human perception of colors; by converting the RGB color space to the HSV color space, it is possible to more conveniently analyze the continuity and changes of colors, thereby detecting delamination defects; for example, when there are delamination defects, it may cause changes in the hue , saturation or lightness of the color.

[0171] Step 306 assumes that the value of a certain pixel point in the HSV color space is , , and the value of its adjacent pixel point in the HSV color space is , , ;

[0172] The formula for calculating the hue difference is as follows:

[0173] ;

[0174] The formula for calculating the saturation difference is as follows:

[0175] ;

[0176] The formula for calculating the lightness difference is as follows:

[0177] ;

[0178] Based on the hue differences of several pairs of pixels , saturation differences and lightness differences calculate the average value of hue differences , average value of saturation differences and average value of lightness differences , and the formula is as follows:

[0179] ;

[0180] The method for calculating the proportion of color abnormal pixels is as follows:

[0181] Preset color discontinuity determination, hue threshold , saturation threshold , lightness threshold , when , , , if any one of the three is satisfied, it is determined that this pair of pixels is color discontinuous in the corresponding attribute;

[0182] Hue threshold , saturation threshold , lightness threshold are set according to the accuracy requirements of the color;

[0183] It should be noted that similar to the grayscale value difference, by calculating the differences of adjacent pixels in each attribute of the HSV color space and comparing them with the corresponding thresholds to determine whether the color is continuous; when the difference exceeds the threshold, it indicates that a large change has occurred in the color in this attribute, which may be caused by color changes due to material or structural changes resulting from layering defects.

[0184] Step 307: Obtain the number of color abnormal pixels , and calculate the proportion of color abnormal pixels through the number of color abnormal pixels The formula is as follows:

[0185] ;

[0186] Among them: the proportion of color abnormal pixels It is usually widely used in the fields of image quality assessment, image processing, computer vision, etc. to quantify the degree to which the color performance in an image does not meet expectations. Its origin is mainly to meet the need for quantitative analysis of image quality, so as to better evaluate the accuracy and stability of the image and perform operations such as image optimization and restoration; this ratio represents the proportion of pixel points with abnormal color conditions such as color distortion, color cast, color oversaturation or undersaturation in the entire image pixel points in an image. The higher the ratio, the more serious the color abnormality in the image, and the lower the image quality may be;

[0187] Step 308: Compare the average gray value difference with the gray value change abnormal determination threshold When the average gray value difference is determined to be abnormal;

[0188] Set the gray value abnormal ratio threshold ;

[0189] Compare the proportion of gray value abnormal pixel points with the gray value abnormal ratio threshold When the proportion of gray value abnormal pixel points is determined to be abnormal;

[0190] The gray value abnormal ratio threshold is set according to the accuracy requirement of the gray value;

[0191] When one of the average gray value difference and the proportion of gray value abnormal pixel points is abnormal, it is determined that the gray value abnormal data is abnormal;

[0192] The specific method for determining whether the color abnormal data is abnormal is as follows:

[0193] Compare the average hue difference with the hue threshold the average saturation difference with the saturation threshold the average lightness difference with the lightness threshold When , , if one of the three cases is abnormal, it is determined that the color of this pixel pair is discontinuous in the corresponding attribute;

[0194] Set the threshold of the color abnormal ratio ;

[0195] Compare the ratio of color - abnormal pixel points with the threshold of the color - abnormal ratio When the ratio of color - abnormal pixel points is abnormal;

[0196] The threshold of the color - abnormal ratio is set according to the requirement for color accuracy;

[0197] When one of the average hue difference , average saturation difference , average lightness difference and the ratio of color - abnormal pixel points is abnormal, determine whether the color - abnormal data is abnormal;

[0198] When one of the abnormal grayscale - value data and the color - abnormal data is abnormal, it is determined that the integrated - circuit package has a delamination defect. When the integrated - circuit package has a delamination defect, a warning is issued.

[0199] By quickly analyzing the internal - structure image, potential delamination defects can be detected in a timely manner without damaging the package, reducing the detection cost and time, and improving production efficiency.

[0200] Step 4: According to the abnormal grayscale - value data and color - abnormal data of multiple integrated - circuit packages, classify the quality level of each integrated - circuit package and generate a production - quality - level report.

[0201] Step 4 includes the following content:

[0202] Step 401: Obtain the average grayscale - value difference , the ratio of grayscale - abnormal pixel points , average hue difference , average saturation difference , average lightness difference and the ratio of color - abnormal pixel points of each normal product, and calculate the quality score , and the formula is as follows:

[0203] ;

[0204] It should be noted that by comprehensively considering multiple parameters related to image quality and assigning different weights to reflect their different degrees of influence on product quality, a comprehensive quality score is obtained; this formula performs weighted summation on multiple image - feature parameters related to product quality to obtain a comprehensive value to represent the quality of the product. The weight of each parameter reflects its relative importance in quality evaluation.

[0205] Step 402: Set the quality score threshold for first-class products as , When it is, it is a first-class product; the quality score threshold for first-class products is , , it is a first-class product; the quality score threshold for qualified products is , When it is, it is a qualified product;

[0206] The quality score threshold for first-class products is , the quality score threshold for first-class products is and the quality score threshold for qualified products is which are set according to the quality scores of different products;

[0207] Statistical total number of products of integrated circuit packages , number of first-class products , number of first-class products , number of qualified products and number of unqualified products ; Calculate the proportion of first-class products , proportion of first-class products , proportion of qualified products and proportion of unqualified products The formulas are as follows:

[0208] ;

[0209] Preset unqualified product threshold , when the unqualified proportion > unqualified product threshold , give an early warning;

[0210] Unqualified product threshold is set by according to the quality scores of different products;

[0211] According to the proportion of first-class products , proportion of first-class products , proportion of qualified products and proportion of unqualified products generate a production quality grade report;

[0212] It can classify the quality grades of multiple packages and generate a detailed report, which is helpful for quality monitoring and improvement in the production process, thereby improving the overall quality of the products

[0213] On the other hand, the present invention also discloses a hierarchical defect detection system based on integrated circuit packaging for implementing the above hierarchical defect detection method, including:

[0214] Data acquisition and frequency calculation module: Obtain the material data and defect size data of the integrated circuit package, and calculate the matching ultrasonic frequency range according to the material data and defect size data;

[0215] Ultrasonic emission module, which emits ultrasonic waves that conform to the ultrasonic frequency range to the integrated circuit package through an ultrasonic emission device;

[0216] Signal reception and image construction module, which receives ultrasonic signals and constructs an internal structure image of the integrated circuit package;

[0217] Defect analysis module, which analyzes the internal structure image, determines whether there are abnormalities in the gray values of adjacent pixel points in the internal structure image and whether there are continuity abnormalities in the colors of adjacent pixel points, generates gray value abnormality data and color abnormality data respectively, and determines whether there is a delamination defect in the integrated circuit package according to the gray value abnormality data and color abnormality data;

[0218] Quality grading and report generation module: According to the gray value abnormality data and color abnormality data of multiple integrated circuit packages, conduct quality grade classification for each integrated circuit package and generate a production quality grade report;

[0219] Each module has a clear division of labor and works collaboratively, reducing redundant steps and time waste in the detection process and improving the overall detection efficiency

[0220] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution.

[0221] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0222] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all of them should be covered by the protection scope of the present application.

Claims

1. A delamination defect detection method based on integrated circuit packaging, characterized in that: The following steps are involved: Acquire material data and defect size data of the integrated circuit package; calculate the ultrasonic frequency range matching the material data and defect size data; Transmitting ultrasonic waves that conform to the ultrasonic frequency range to the integrated circuit package through an ultrasonic transmitting device; receiving ultrasonic signals and constructing an internal structure image of the integrated circuit package; Analyze the internal structure image, determine whether there is an abnormality in the grayscale values ​​of adjacent pixels of the internal structure image and whether there is an abnormality in the continuity of the colors of adjacent pixels, generate grayscale value abnormality data and color abnormality data respectively, and determine whether there is a delamination defect in the integrated circuit package based on the grayscale value abnormality data and the color abnormality data; According to the gray value abnormality data and color abnormality data of multiple integrated circuit packages, each integrated circuit package is classified into a quality grade, and a production quality grade report is generated.

2. The method for detecting delamination defects based on integrated circuit packaging according to claim 1, characterized in that: The method for calculating the ultrasonic frequency range is as follows: Select the wave velocity interval data that matches the material data ; Calculate wavelength data based on defect size data , specifically: Defect size data includes at least the minimum delamination defect size , according to the minimum delamination defect size Calculating wavelength data , the formula is as follows: ; According to the wave speed interval data and wavelength data Calculate the ultrasonic frequency range , the formula is as follows: 。 3. The method for detecting delamination defects based on integrated circuit packaging according to claim 2, characterized in that: The method to construct the internal structure image is as follows: Obtain the intensity of ultrasonic waves reflected at different measurement positions , through the ultrasonic intensity Calculate the grayscale value at different measurement positions , based on the following formula: ; in: and is a constant, and the ultrasonic intensity is Mapping to grayscale values ; When constructing the image, different measurement locations are mapped to pixel locations in the image. The measurement position corresponds to the pixels; the calculated gray value Assign to this pixel.

4. The method for detecting delamination defects based on integrated circuit packaging according to claim 3, characterized in that: Calculate the received ultrasonic signal strength The method is as follows: Get ultrasonic transmission power The distance of ultrasonic wave propagation , through ultrasonic transmission power The distance of ultrasonic wave propagation Calculate the received ultrasonic signal strength , the formula is as follows: 。 5. The method for detecting delamination defects based on integrated circuit packaging according to claim 4, characterized in that: Calculate ultrasonic transmission power The method is as follows: Get medium density , speed of sound , vibration area , angular frequency and amplitude , through the medium density , speed of sound , vibration area , angular frequency and amplitude Calculate ultrasonic transmission power , the formula is as follows: 。 6. The method for detecting delamination defects based on integrated circuit packaging according to claim 5, characterized in that: Gray value abnormal data includes: average gray value difference and the ratio of abnormal pixels in grayscale value ; Among them, the average gray value difference The calculation method is as follows: Get several pixels through the internal structure image The gray value is and its adjacent pixels The gray value is , through the pixel The gray value is and its adjacent pixels The gray value is Calculate the gray value difference between adjacent pixels , based on the following formula: ; The gray value difference of adjacent pixels , calculate the mean gray value difference , based on the following formula: ; in: It is For the gray value difference of adjacent pixels, The value range is 1, 2, 3... , is the number of adjacent pixel pairs selected , the value is a positive integer; The ratio of abnormal gray value pixels The calculation method is as follows: Set the grayscale change abnormality judgment threshold ; Gray value change abnormality judgment threshold The calculation method is as follows: ; in: is the mean gray value difference The coefficient of When , it is determined that the grayscale change of this pixel pair is abnormal; Get the number of grayscale value abnormal pixel pairs , by the number of abnormal gray value pixel pairs Calculate the ratio of abnormal grayscale pixels The formula is as follows: ; in: is the total number of adjacent pixel pairs.

7. The method for detecting delamination defects based on integrated circuit packaging according to claim 6, characterized in that: Color anomaly data includes: average hue difference , Saturation difference average , average value of brightness difference and color abnormal pixel ratio ; Among them, the average value of hue difference , Saturation difference average and the average value of brightness difference The calculation method is as follows: The method for determining whether there is continuity anomaly in the colors of adjacent pixels is as follows: set up are the red, green, and blue color component values ​​of a pixel in the RGB color space, and their value ranges are [0,255]. ; Represents the maximum value of the red, green, and blue color components of the pixel. Represents the minimum value of the red, green, and blue color components of the pixel, and then calculates the hue : when hour, ; when and hour, ; when and hour, ; when hour, ; when hour, ; Calculate saturation : when hour, ; when hour, ; Calculate brightness ; Set pixel The value in HSV color space is , , Its neighboring pixels The value in HSV color space is , , ; Calculate hue difference The formula is as follows: ; Calculate saturation difference The formula is as follows: ; Calculate brightness difference The formula is as follows: ; Based on the hue difference of several pixel pairs , Saturation Difference and brightness difference Calculate the average hue difference , Saturation difference average and the average value of brightness difference , the formula is as follows: ; in: is the number of adjacent pixel pairs selected, It is For the hue difference of adjacent pixels, The value range of is 1, 2, 3...d, d is a positive integer, It is The saturation difference of adjacent pixels, It is The brightness difference of adjacent pixels; Calculate the ratio of color abnormal pixels The method is as follows: Preset color discontinuity judgment, hue threshold , Saturation threshold , brightness threshold ,when , , When one of the three conditions is satisfied, it is determined that the color of this pixel pair is discontinuous in the corresponding attribute, where: Is the hue difference The weight coefficient ranges from 0.2 to 0.

3. is the saturation difference Weight coefficient, value range 0.3~0.4, Is the brightness difference The weight coefficient ranges from 0.3 to 0.

4. + + =1; Get the number of color abnormal pixels , by the number of color abnormal pixels Calculate the ratio of color abnormal pixels The formula is as follows: 。 8. The method for detecting delamination defects based on integrated circuit packaging according to claim 7, characterized in that: The method for determining whether an integrated circuit package has a delamination defect is: The specific method for determining whether the gray value abnormal data is abnormal is: The mean gray value difference Gray value change abnormality judgment threshold For comparison, When the average gray value difference is determined abnormal; Set the gray value abnormal ratio threshold ; The ratio of abnormal grayscale pixels Gray value abnormal ratio threshold Compare, when When the gray value is abnormal, the proportion of pixels is determined abnormal; When the mean gray value difference Ratio of abnormal pixels to gray value If one of the items is abnormal, it is determined that the gray value abnormal data is abnormal; The specific method for determining whether color abnormality data is abnormal is: Average the hue differences With hue threshold , Saturation difference average With saturation threshold , average value of brightness difference With brightness threshold For comparison, , , When , one of the three situations is abnormal, and it is determined that the color of this pixel pair is discontinuous in the corresponding attribute; Set the color anomaly ratio threshold ; The ratio of abnormal color pixels Threshold value for color anomaly ratio Compare, when When the color abnormality pixel ratio is determined abnormal; When the average hue difference , Saturation difference average , average value of brightness difference and color abnormal pixel ratio If one of the items is abnormal, it is determined whether the color abnormality data is abnormal; When one of the gray value abnormal data and the color abnormal data is abnormal, it is determined that the integrated circuit package has a delamination defect, and when the integrated circuit package has a delamination defect, an early warning is issued.

9. The method for detecting delamination defects based on integrated circuit packaging according to claim 8, characterized in that: To generate a production quality grade report: Get the average gray value difference of each normal product , The ratio of abnormal gray value pixels , average hue difference , Saturation difference average , average value of brightness difference and color abnormal pixel ratio , calculate the quality score , the formula is as follows: ; in: is the mean gray value difference The weight coefficient ranges from 0.1 to 0.

2. is the ratio of abnormal grayscale pixels The weight coefficient ranges from 0.2 to 0.

3. is the average hue difference The weight coefficient ranges from 0.1 to 0.

2. is the average saturation difference The weight coefficient is 0.2 to 0.

3. is the average value of brightness difference The weight coefficient ranges from 0.1 to 0.

2. is the ratio of color abnormal pixels The weight coefficient ranges from 0.2 to 0.3, and + + + + + =1; Set the quality score threshold of superior products to , When , it is a superior product; the quality score threshold of the first-class product is , , is a first-class product; the quality score threshold of qualified products is , When it is, it is a qualified product; Count the total number of integrated circuit packaging products 、Quantity of superior products 、First-class product quantity , Quantity of qualified products and the number of defective products ; Calculate the proportion of superior products , first-class product ratio , Proportion of qualified products and the proportion of defective products The formula is as follows: ; Preset threshold for non-conforming products , when the unqualified ratio >Unqualified product threshold When the alarm is sounded; According to the proportion of superior products , first-class product ratio , Proportion of qualified products and the proportion of defective products Generate production quality grade reports.

10. A layered defect detection system based on integrated circuit packaging, characterized in that: Data acquisition and frequency calculation module: obtains the material data and defect size data of the integrated circuit package, and calculates the matching ultrasonic frequency range according to the material data and defect size data; An ultrasonic transmitting module transmits ultrasonic waves conforming to the ultrasonic frequency range to the integrated circuit package through an ultrasonic transmitting device; A signal receiving and image building module receives ultrasonic signals and builds an internal structure image of the integrated circuit package; The defect analysis module analyzes the internal structure image, determines whether there is an abnormality in the grayscale values ​​of adjacent pixels in the internal structure image and whether there is an abnormality in the continuity of the colors of adjacent pixels, generates grayscale value abnormality data and color abnormality data respectively, and determines whether there is a delamination defect in the integrated circuit package based on the grayscale value abnormality data and the color abnormality data; Quality grading and report generation module: According to the gray value abnormality data and color abnormality data of multiple integrated circuit packages, each integrated circuit package is classified into a quality grade and a production quality grade report is generated.

Images