This paper treats the intelligent inspection system for particle suspension by using a neural network. The automated inspection system for particle suspension by an image processing has almost not been reported in the past, because of the complexity of image processing of water with particles. If the features of image processing data of particle suspension are recognized by a vision system and a neural network, the degree of surface contamination before and after washing processes will be

more »

... ied in relation to the water analyses, and also the deterioration process of many kinds of fluid, such as lubricants and finishing compounds will be cleared. The objectives of the paper are to analyze the water image by using the texture analyses, to investigate the limitation of image recognition for concentration of particles in water, and to construct the neural network system which' is able to estimate the concentration of particle suspension. It is found that the features of image data of the liquid containing particles can be decided by the texture analyses. As the results, it is possible to construct the neural network system which is composed of the features of water image as input data and the degree of particles concentration as output data.