Detecting algorithm for eccentrically-inserting defect of LED luminous chip

Abstract

The invention discloses a detecting algorithm for an eccentrically-inserting defect of an LED luminous chip. A colorful image I of the LED luminous chip is collected; the colorful image I of the LED luminous chip which is to be detected is read by a computer; grayscale processing and de-noising processing are conducted on the colorful image I, and an image I1 is obtained; binarization processing is conducted on the image I1, and an image I2 is obtained after the binarization processing; refining processing is conducted on the image I2 obtained after the binarization processing, and an image I3 is obtained after the refining processing; linear fitting is conducted on the image I3 obtained after the refining processing, and a fit straight line is obtained; whether the eccentrically-inserting defect of the LED luminous chip exists is judged according to the fit straight line. The detecting algorithm for the eccentrically-inserting defect of the LED luminous chip can carry out real-time detection on the eccentrically-inserting defect of LED luminous chips in a production line, and is efficient and accurate and can be applied to real-time on-site operation.

Description

technical field [0001] The invention belongs to the technical field of computer image processing, in particular to a detection algorithm for insertion deviation defects of LED light-emitting chips. Background technique [0002] In recent years, under the condition of the full popularity of fully automatic packaging lines, on-line inspection of the packaging quality of LED light-emitting chips during the packaging production process has become an inevitable requirement to improve the packaging level and ensure the packaging quality. However, due to the small size of LED chips, high packaging process requirements, and fast production line speed, it is difficult to perform real-time quality inspection during the packaging process. Due to the limitations of the existing sorting technology, only a small amount of packaged and sorted LED products can meet the requirements of a certain customer (such as a high-end LED display manufacturer), and most of the rest will become inventor...

AI Technical Summary

Problems solved by technology

Since the luminous uniformity and brightness performance of the LED are mainly determined by the position of the light-emitting chip in the sealant, and the light-emitting chip usually has insertion deviation defects in the sealant, if the insertion deviation value is at the micron level, it will not significantly affect the LED. The average spectral intensity and the complete spectral irradiance will have an impact, but it will have a significant impact on the uniformity of the light intensity distribution of the LED, thereby affecting the optical performance of the LED

[0004] Existing sorting equipment such as spectrometers and color separation machines can only detect and classify from the average spectral intensity and complete spectral irradiance of LED light-emitting chips after power-on, and cannot distinguish the uniformity of light intensity distribution of LED lamp beads. Therefore, it is impossible to identify subtle insertion deviation defects using these devices, so that the color temperature and color difference and luminous uniformity of the LED display cannot be guaranteed (that is, the dispersion of the directional characteristics)

In addition, existing sorting equipment such as integrating spheres and photomultiplier tubes used to detect the uniformity of LED light intensity distribution cannot achieve online rapid and continuous sorting of LED light intensity distribution uniformity due to the limitation of detection principles. type sorting is more difficult to achieve

Images