This research explores the automated detection of surface blemishes that fall across two different background textures in a light-emitting-diode (LED) chip. Water-drop blemishes, commonly found on chip surfaces, impair the appearance of LEDs as well as their functionality and security. Automated inspection of a water-drop blemish is difficult, because the blemish has a semi-opaque appearance and a low intensity contrast with the rough exterior of the LED chip. Moreover, the blemish may fall across two different background textures, which further increases the difficulties of defect detection. We first use the one-level Haar wavelet transform to decompose a chip image and extract four wavelet characteristics. Then, the T2 statistic of multivariate statistical analysis is applied to integrate the multiple wavelet characteristics. Thus, the wavelet-based T2 approach judges the existence of water-drop blemishes. Finally, we compare the defect detection performance among the wavelet-based T2 method and traditional methods. Experimental results show that the proposed method achieves a 95.8% probability of accurately detecting the existence of water-drop blemishes, and an approximate 92.6% probability of correctly segmenting their regions.