CN-122016815-A - Method, device and system for detecting omnibearing defect of light-emitting element based on multi-view imaging

CN122016815ACN 122016815 ACN122016815 ACN 122016815ACN-122016815-A

Abstract

The application provides a method, a device and a system for detecting the omnibearing defect of a luminous element based on multi-view imaging, wherein the method comprises the steps of arranging a display module to be detected on a defect detection jig, and acquiring detection data of the luminous element in the display module and a plurality of spectrum images with different view angles; and carrying out pretreatment and feature extraction on each spectrum image to obtain feature data, and judging according to the feature data and the detection data to obtain a detection result. The omnibearing defect detection method for the luminous element based on the multi-view imaging adopts a plurality of spectrum images of different view angles of the luminous element, combines detection data and characteristic data to judge, realizes one-time omnibearing defect detection, covers a plurality of key quality indexes such as appearance, photoelectric performance, uniformity and the like, enriches detection dimensions, improves detection efficiency, reduces omission ratio, realizes omnibearing no-dead-angle acquisition of spectrum images and eliminates detection dead areas of single view angles.

Inventors

LIANG HONG
ZHANG JUN
GONG DEWANG
YAN HAIBO
ZHONG WEIXIANG

Assignees

上海珺远光电科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260310

Claims (10)

1. The omnibearing defect detection method for the light-emitting element based on the multi-view imaging is characterized by comprising the following steps of: Setting a display module to be detected on a defect detection jig, and acquiring detection data of a light-emitting element in the display module and a plurality of spectrum images with different visual angles; Preprocessing each spectrum image and extracting features to obtain feature data; And judging according to the characteristic data and the detection data to obtain a detection result.
2. The method for detecting the omnibearing defect of the light-emitting element based on multi-view imaging according to claim 1, wherein the characteristic data comprises appearance data, an image average brightness value, an image gray scale standard deviation and a color coordinate, the detection data comprises forward current, reverse current, luminous flux and luminous intensity, the judgment is carried out according to the characteristic data and the detection data, and the detection result comprises: If any one or more of scratches, black points, impurities and bubbles exist in the appearance data, the detection result is that the appearance of the light-emitting element is disqualified; If the luminous intensity and/or the luminous flux are not in the set allowable range, the detection result is that the photoelectric performance of the light-emitting element is unqualified; if the reverse current is larger than the forward current, the detection result is that the light-emitting element leaks electricity; if the color coordinates are not in the target setting area, the detection result is that the color of the light-emitting element is unqualified; And if the average brightness value of the image is not greater than a set brightness lower limit threshold value or the standard deviation of the gray scale of the image is greater than a set gray scale threshold value, the detection result is that the gray scale distribution of the light emitting surface of the light emitting element is uneven.
3. The method for detecting the omnibearing defect of the light-emitting element based on multi-view imaging according to claim 1, further comprising the step of obtaining a transmission image of the light-emitting element in the display module, wherein if gold wires are broken and/or die bonding cavities exist in the transmission image, the detection result is that the defect exists in the light-emitting element.
4. A method for detecting an omnibearing defect of a light emitting element based on multi-view imaging according to any one of claims 1 to 3, wherein preprocessing and feature extraction are performed on each of the spectral images to obtain feature data comprising: carrying out graying, noise reduction and image enhancement treatment on each spectrum image to obtain a treatment image corresponding to each spectrum image; positioning all the processed images and segmenting the region of interest to obtain picture data containing a luminous region, an electrode region and the edge of a package body; and extracting gray features, RGB color components, geometric features and texture features of all the picture data to obtain feature data.
5. The method for detecting the omnibearing defect of the light emitting element based on the multi-view imaging according to any one of claims 1 to 3, wherein the step of obtaining the multiple spectral images of the light emitting element in different view angles in the display module comprises the steps of obtaining the multiple spectral images of the light emitting element in different view angles from the upper side, the side and the lower side of the light emitting element according to a visible light mode or a specific wave band mode by adopting an imaging device, or the step of obtaining the detection data of the light emitting element in the display module comprises the steps of respectively applying rated working current to the light emitting element from the forward direction and the reverse direction to obtain forward current and reverse current, and the step of applying a standard light source to the light emitting element to obtain luminous flux and luminous intensity.
6. The omnibearing defect detection device for the light-emitting element based on the multi-view imaging is characterized by comprising a data acquisition unit, a data processing and feature extraction unit and a judgment detection unit; the data acquisition unit is used for arranging a display module to be detected on the defect detection jig and acquiring detection data of a light-emitting element in the display module and a plurality of spectrum images with different visual angles; The data processing and feature extraction unit is used for preprocessing each spectrum image and extracting features to obtain feature data; and the judging and detecting unit is used for judging according to the characteristic data and the detection data to obtain a detection result.
7. The multi-view imaging-based light emitting element omnibearing defect detection apparatus according to claim 6, wherein the characteristic data includes appearance data, an image average luminance value, an image gray scale standard deviation, and color coordinates, the detection data includes forward current, reverse current, luminous flux, and luminous intensity, and the judgment detection unit includes; If any one or more of scratches, black points, impurities and bubbles exist in the appearance data, the detection result is that the appearance of the light-emitting element is disqualified; If the luminous intensity and/or the luminous flux are not in the set allowable range, the detection result is that the photoelectric performance of the light-emitting element is unqualified; if the reverse current is larger than the forward current, the detection result is that the light-emitting element leaks electricity; if the color coordinates are not in the target setting area, the detection result is that the color of the light-emitting element is unqualified; And if the average brightness value of the image is not greater than a set brightness lower limit threshold value or the standard deviation of the gray scale of the image is greater than a set gray scale threshold value, the detection result is that the gray scale distribution of the light emitting surface of the light emitting element is uneven.
8. The device for detecting the omnibearing defect of the light-emitting element based on multi-view imaging according to claim 6, wherein the data acquisition unit is further used for acquiring a transmission image of the light-emitting element in the display module, and the judging and detecting unit is further used for judging whether gold wires are broken and/or die bonding holes exist in the transmission image, and the detection result is that the defect exists in the light-emitting element.
9. The multi-view imaging-based light emitting element omnibearing defect detection device according to claim 6, wherein the data processing and feature extraction unit comprises a preprocessing subunit, a segmentation subunit, and a feature extraction subunit; The preprocessing subunit is used for carrying out graying, noise reduction and image enhancement processing on each spectrum image to obtain a processed image corresponding to each spectrum image; The segmentation subunit is used for positioning all the processed images and segmenting the region of interest to obtain picture data containing a luminous region, an electrode region and the edge of the package body; the feature extraction subunit is configured to extract gray features, RGB color components, geometric features, and texture features of all the image data to obtain feature data.
10. The system is characterized by comprising a defect detection jig and an industrial personal computer, wherein a detection station for placing the light-emitting element in the display module is arranged on the defect detection jig, and the industrial personal computer automatically detects defects of the light-emitting element in the detection station according to the method for detecting the omnibearing defect of the light-emitting element based on the multi-view imaging as set forth in any one of claims 1 to 5.

Description

Method, device and system for detecting omnibearing defect of light-emitting element based on multi-view imaging Technical Field The application belongs to the technical field of illumination defect detection, and particularly relates to a method, a device and a system for detecting omnibearing defects of a light-emitting element based on multi-view imaging. Background The existing method for detecting defects of the LED lamp bead display module in the traditional manual visual inspection identification image comprises the steps of adopting CCD equipment to acquire images, adopting a manual visual inspection mode, a single visual angle detection mode and a point scanning detection mode, wherein the detection mode is low in efficiency, strong in subjectivity, easy to fatigue, different in standard and easy to miss inspection, and cannot meet the efficiency requirement of a modern production line. The single visual angle detection mode refers to shooting the light emitting surface of the LED lamp bead only from the position right above the LED lamp bead display module, and the problems of abnormal side light emission or uneven light color distribution caused by side surfaces (such as scratch and glue shortage of packaging colloid), electrode oxidation, internal gold wire breakage, chip die bonding defects and the like of the lamp bead cannot be detected. The single visual angle detection mode has a detection blind area and high omission ratio. The point scanning detection mode has low efficiency and cannot be used for rapid full detection. Disclosure of Invention An embodiment of the application provides a method, a device and a system for detecting omnibearing defects of a light-emitting element based on multi-view imaging, which are used for solving the problems that the existing defect detection mode of an LED lamp bead display module is incomplete in azimuth, low in detection efficiency and easy to miss detection. In a first aspect, an embodiment of the present application provides a method for detecting an omnibearing defect of a light emitting element based on multi-view imaging, including the steps of: Setting a display module to be detected on a defect detection jig, and acquiring detection data of a light-emitting element in the display module and a plurality of spectrum images with different visual angles; Preprocessing each spectrum image and extracting features to obtain feature data; And judging according to the characteristic data and the detection data to obtain a detection result. Optionally, the feature data includes appearance data, an average brightness value of an image, a standard deviation of gray scale of the image, and color coordinates, the detection data includes forward current, reverse current, luminous flux, and luminous intensity, and the determining according to the feature data and the detection data, and the obtaining the detection result includes: If any one or more of scratches, black points, impurities and bubbles exist in the appearance data, the detection result is that the appearance of the light-emitting element is disqualified; If the luminous intensity and/or the luminous flux are not in the set allowable range, the detection result is that the photoelectric performance of the light-emitting element is unqualified; if the reverse current is larger than the forward current, the detection result is that the light-emitting element leaks electricity; if the color coordinates are not in the target setting area, the detection result is that the color of the light-emitting element is unqualified; And if the average brightness value of the image is not greater than a set brightness lower limit threshold value or the standard deviation of the gray scale of the image is greater than a set gray scale threshold value, the detection result is that the gray scale distribution of the light emitting surface of the light emitting element is uneven. Optionally, the omnibearing defect detection method for the luminous element based on multi-view imaging further comprises the steps of obtaining a transmission image of the luminous element in the display module, and if gold wires are broken and/or die bonding cavities exist in the transmission image, detecting that defects exist in the luminous element. Optionally, preprocessing and feature extraction are performed on each spectrum image, and obtaining feature data includes: carrying out graying, noise reduction and image enhancement treatment on each spectrum image to obtain a treatment image corresponding to each spectrum image; positioning all the processed images and segmenting the region of interest to obtain picture data containing a luminous region, an electrode region and the edge of a package body; and extracting gray features, RGB color components, geometric features and texture features of all the picture data to obtain feature data. Optionally, acquiring the spectrum images of the light emitting elements in the display module