Finish chinks are regarded as the most critical defects of recycled bottles adopted in beverage production. This paper presents a bottle finish inspection system utilizing state-of-art machine vision technologies to implement high-speed online inspection. The system architecture is illustrated in detail. An algorithm derived from Hough transform is applied for real-time determination of area of interest. As for detection, two artificial neural networks are used for low-level inspection and high-level judgment respectively. A prototype is developed and experimental results demonstrate the feasibility of the inspection system. The proposed algorithms are proved to be efficient in automatic finish inspection.