A case study: passive component inspection using a 1D wavelet transform

Abstract

This paper exploits a wavelet-based scheme to inspect the surface defects and basic dimensions of 0805 Multi-layer Ceramic Chip capacitors (MLCC) using machine vision. The image of a passive component is initially processed to show only two solder plates (terminations). Then, the covariance matrix eigenvector for each boundary point generates the 1D h-p representation to describe the angle variations at the boundaries of each termination. The 1D h-p representation is further decomposed directly by a one-dimensional wavelet transform (1D WT). Since a single corner (an intersection of two boundary edges) and the jag corners (the surface defects) are local deviations on termination boundaries, their locations are represented as intensive and have a highly fluctuated wavelet energy at the 1st detail scale. Concerning a 0805 MLCC type of passive component, the surface defects and single corners on termination boundaries can be captured by appropriate thresholds (e.g., in wavelet energy). The basic dimensions of a passive component are simply the direct distance between located single corners. Real 0805 MLCCs are testing samples to evaluate the performance of the proposed approach. Experimental results show that the proposed method achieves a precise identification of the surface defects and measurements for basic dimensions. The proposed approach is invariant with respect to the orientation, easy to implement and free from a primitive-matching process. Therefore, it is especially suitable for various types of passive components that are similar to the 0805 MLCC in small-batch production.

Tiny surface defect inspection of electronic passive components using discrete cosine transform decomposition and cumulative sum techniques

“…As to inspecting surface flaws of electronic components, some machine vision systems have been applied to PCBs [9,17], Ball Grid Array (BGA) substrate conducting paths [10], and capacitor chips of passive components [3,18]. Lu and Tsai [6], and Lu and Tsai [7] use singular value decomposition to inspect surface defects of Thin Film TransistorLiquid Crystal Display (TFT-LCD) panels.…”

“…Surface defects affect not only the appearances of passive components but also their functionality, efficiency and stability. Large and obvious surface defects such as indents, scraps and scratches are usually inspected by automated visual inspection systems [3][4][5][6][7][8]. But tiny surface flaws such as dust, cavities and pinholes are very difficult to detect because of their extremely small sizes.…”

“…More than 80% of the parts in a motherboard are composed of passive components [1,2]. Also, the sizes of passive components have shrunk to 5 · 2.5 · 2.5 mm for sophisticated products [3].…”

Revision of Using Eigenvalues of Covariance Matrices in Boundary-Based Corner Detection

“…Since then, many studies have applied the small eigenvalues to detect corners directly [15], [16], [18], [21]. In addition, some others studies were also inspired by Tsai et al's small eigenvalues approach [6], [7], [13], [14], [17], [19], [22], [23]. Although Sossa Azuela et al [20] commented that Tsai et al's method was the best one among the methods they had tested, Guru et al [8] later discovered that Tsai et al's method may also detect unwanted spurious corners.…”

Tombstone Initiation Model for Small Form-Factor Surface Mount Passives

“…For example, any defects or disturbances in the lead-free metallization of capacitor terminals (typically a low-meltingpoint metal, such as Sn [6], [7]) can become significant if the terminals shrink with the capacitor FF [8]. Such disturbances are not an issue for large FF capacitors due to their relative size compared to the overall termination.…”