The automatic detection system of flexible printed circuit board’s defect in computer equipment based on picture processing and pattern recognition

Abstract

Purpose – The purpose of this paper is to make the test of flexible printed circuit (FPC) board be more efficient and accurate. Design/methodology/approach – This paper aims to segment an image through applying the technology of computer pattern recognition. Findings – The results indicate that the automatic detection system of FPC board’s defect has high efficiency and accuracy in accordance with the requirements of detection. Originality/value – This paper managed image segmentation through applying the technology of computer pattern recognition.

Flexible Printed Circuit Fracture Detection Based on Hypothesis Testing Strategy

“…Reference [16] calculates the number of connected domains in the image to analyze the number of circuits statistically, so as to determine whether there is any break defect. In [17], the number of the track borders in each small local area is counted, and then the defect is detected qualitatively according to the number of borders perpendicular to the track extension direction. (2) Another category is for the precise location of the defect.…”

“…[B LSU (M i )] is the length of the basic track structure unit of the morpheme corresponding to M i in the corpus . In (17), k 16 represents the sum of the number of the morphemes whose values are 1 or 6 in…”

“…To illustrate the classification performance of the algorithm we proposed, it is compared with the methods described in [9], [13], [15], [17], and [20]. The reason for choosing the methods in the above six references is that: 1) The result of the method or its certain step can obtain the exact location of the fracture.…”

“…In the compared methods, [9] is the representative of the direct comparison method (DCM), [13] is the representative of indirect comparison method (ICM), and [15] is the representative of feature comparison method (FCM). Reference [17] and [20] are two different rule-based methods, where [20] takes the difference between the track edge and the fitted line as the detection rule, called straight line fitting method (SLFM). And in [17], the edge shape of the local region of the track is used as the detection rule, called local edge configuration method (LECM).…”

“…(2) The rule-based method designs a series of digital image processing rules according to the basic criterion of circuit printing to detect defects. However, a part of these researches can only achieve qualitative inspection [16], [17]. The other parts can realize the location of the fracture, but they only consider the ''simple'' case that the fracture occurs on a single track.…”