Wafer map defect pattern recognition provides useful clues to yield learning. However, most wafer maps have no special spatial patterns and are full of noises, which make pattern recognition difficult. Especially, recognizing scratch and line types of defect patterns is challenging for process and test engineers. It takes a lot of manpower to identify such patterns, as hidden defective dies may exist on the scratch contour and become discontinuity points. Hidden scratch defective dies may suffer from latent and leakage faults, which usually deteriorate quickly and need to be screened by burn-in test to improve quality. A possible solution is to locate the obscure defective dies in scratch patterns and mark them as faulty. As a result, the quality and reliability of products is significantly improved and cost of final test is reduced. In this article, we propose a systematic methodology to search for potential hidden scratch/line defects in wafers. A five-phase method is developed to enhance wafer maps such that automatic hidden scratch defect pattern recognition can be carried out with high accuracy. Experimental results show the proposed method achieves higher than 89% recognition rate for scratch/line patterns, and higher than 94% for all common wafer defect pattern types.

中文翻译：

---

隐藏晶圆划痕缺陷投影以进行诊断和质量增强

晶圆图缺陷模式识别为良率学习提供了有用的线索。然而，大多数晶片图没有特殊的空间图案并且充满了噪声，这使得图案识别变得困难。尤其是，识别缺陷图案的划痕和线条类型对于过程和测试工程师而言是一项挑战。识别此类图案需要大量人力，因为隐藏的有缺陷的模具可能会存在于刮擦轮廓上并成为不连续点。隐藏的刮擦缺陷模具可能会遭受潜在和漏电故障的影响，这些故障通常会迅速恶化，因此需要通过老化测试进行筛选以提高质量。一种可能的解决方案是以划痕的方式定位模糊的缺陷管芯并将其标记为有缺陷的。结果，显着提高了产品的质量和可靠性，并降低了最终测试的成本。在本文中，我们提出了一种系统的方法来搜索晶圆中潜在的潜在划痕/线条缺陷。开发了一种五阶段方法来增强晶片图，以便可以高精度执行自动隐藏的划痕缺陷图案识别。实验结果表明，该方法对划痕/线条图案的识别率高于89％，对于所有常见晶圆缺陷图案类型的识别率均高于94％。