including low accuracy in detecting defects on small PCBA targets and a high rate of false positives and omissions24. YOLOv5 has been widely adopted for object detection tasks across various fields, and numerous improved models have emerged that could serve as valuable references for enhancing...
Deep learning, particularly CNNs, has been extensively used in the field of steel strip defect detection, gradually replacing traditional machine vision-based methods. These algorithms have gained widespread application in various industrial domains, including the detection of metal surface defects and PCB...
In this study, we aimed to address the primary challenges encountered in industrial integrated circuit (IC) surface defect detection, particularly focusing on the imbalance in information density arising from difficulties in data sample collection. To this end, we have developed a new hybrid architectu...
The scattering signal of a deep-subwavelength defect can be qualitatively expressed by the Rayleigh formula [40], i.e. Idefect=(1+cos2θ2R2)(2πλ)4(|N|2−1|N|2+2)2(d2)6I0, (1) where θ is the incident angle, R is the distance between the observer point and the ...
Defect Detection Percentage: The defect detection percentage (DDP) gives a measure of the testing effectiveness. It is calculated as a ratio of defects found prior to release and after release by customers Formula:DDP=Number of defects at the moment of software version release / Number of defects...
Amongst the existing corrosion assessment methodologies (Barlow's Formula (Eq. (10))), DNV (Eq. 11–13), ASME B31G, and Modified ASME B31G), DNV executes the minimum RMSE value and the maximum RMSE value provided by Barlow's formula. However, in terms of failure pressure prediction, DNV...
As shown in formula (1):(1)f˜=f−cWhere f˜ denotes the defect sample, f is corresponding defect-free sample, c is the missing carbon structure at the defect position. According to the repeatability of ACCC wire texture, the estimation of f˜ can be obtained from other images ...
CTF objects were created with an electron energy of 300 KV, a spherical aberration randomly selected from 1 to 3 μμm and defocus range from −4−4 to 4 nm, and a focal spread of 30 Å The final image was combined using the following formula: GB=GaussianBlur(image,kennelsize),Img...
where N denotes the number of PCB defect types and AP is the area enclosed by the PR curve, the calculation formula is shown in Eq. (16). $$AP={\int }_{0}^{1}P\left(R\right)dR$$ (16) P is the precision, which indicates the probability of being correctly classified in the ...
During the production of metal material, various complex defects may come into being on the surface, together with large amount of background texture information, causing false or missing detection in the process of small defect detection. To resolve those problems, this paper introduces a new mode...