Many visual automatic inspection methods are used in the assembly line of vehicles for improving quality and reducing a production cost. In order to resolve uneasy inspection problems, new useful methods and devices have been developed. A real factory environment has very poor optical conditions. Illuminations are affected by the sun lights passing through the windows of the factories. The floor is continuously vibrating, which could affect the posture of installed cameras and illumination devices. This paper describes how to overcome these problems and suggests the automatic inspection system applicable to the automobile assembly line. It introduces two feature extraction methods and algorithms to detect the opened car in the automobile assembly line. The methods for extracting these features are discussed in this paper with the experimental results.