Research Of Measurement And Statistic Methods For Rock Aggregates Based On Image

In the procedure of making the aggregate for industry and enterprise, it needs to do non-stop online examine and auto-control urgently. Aiming at the control problem of gradating rock aggregates in the asphalt mixtures mixing plant, this paper this paper puts forward the method of collecting, handling and analyzing aggregate images in real time, using the computer-image-examine skill. As a result, it will improve the efficiency and qualityof the production. In this thesis, rock fragments image processing has been researched using the theory of image processing and mathematical morphology based on this researching environment, which can realize the automatic identification of rock boundary and obtain the estimation of characteristic parameters. The corresponding workings are given as follows:1. Aiming at special characteristic of the rock fragments image, the method of the combination Gaussian filter with Wiener filter is put forward in the dissertation, and traditional image segmentation methods are summed up by some experiments.2. Mathematical Morphology Theory is introduced and applied to rock fragments image which is provided a base for research on rock fragments image, for example, image filter, image enhancement, image reconstruction.3. Image segmentation method based on edge growing and side illumination measure is put forward. Side illumination measure can eliminate spurious edges, and an adaptive edge growing arithmetic based on fuzzy logic is advanced to link discontinuous edge in the thinning image, which increases connectedness and correctness, experiment results of the automatic identification of rock boundary by this method are satisfied.4. An image process method of synthetically taking use of boundary information extracting arithmetic, morphological reconstruction and watershed transformation is proposed. Firstly, double threshold binarizating and edge detecting, then distance transforming, where after, morphological reconstruction h-Dome transformation marking the max. areas, lastly, watershed transforming. The experiment result shows that this algorithm can separate the rock fragment aggregate successfully.