No abstract

Using a plastic water hose not only affects water resources, but also affects people's daily lives. Due to over bending or twisting during service times, plastic hoses tend to fold, buckle, or crack. The new developed spiral plastic hose can overcome the above problems, however, how to evaluate its properties is still an industry need. By using machine vision technology to analyze the spiral plastic hose quality such as the spring-back property is a good methodology which can facilitate the trust relationship between the industries and consumers. Three-dimensional machine vision technologies were used to examine five different kinds of plastic hoses, i.e. type A-E. At first, a testing platform is designed and the measuring algorithm is proposed to measure the inner radius of the bent plastic hose. By using image processing technology and the developmental program, the radius information of plastic hoses before and after spring-back are obtained for further analysis. According to experimental results based on the proposed inspection system, Type B plastic hoses behave with no-kink and its spring-back ratio is 3.457, which is the best of the plastic hoses. Plastic hose of Type D behaves kink situation and its spring-back ratio is 1.117 which is the worst. By comparing the experimental results to the manufacturers offering information, a good agreement is obtained. Therefore, this study provides an innovative inspection system to determine plastic hose quality.