| We are faced with a lot of problems in the Image Processing, such as the larger data sets and the slower computing speed. For this reason, the parallel image processing which based on parallel computer or parallel cluster was used to improve the efficiency of image processing. With the emergence of multi-core architecture, multi-core parallel technology has become a new method to increase the speed of image processing. However, the existing parallel design and programming pattern are not enough suitable for multi-core architecture and lack of universality. Thus, it is difficult to extend this application. Currently, the application of multi-core parallel technology in the fields of image processing has attracted extensive attention,the research on multi-core image processing has become a hot research area.Based on the combination of multi-core parallel technology and image processing algorithms, the paper proposes a multi-core image processing parallel design model, which has included the following five steps, parallel analysis, modeling, mapping, debugging &performance evaluation and testing & release. As experience proves, the model is easy to understand and apply. Taken the mosaic of borehole wall expanded images as the research object, the paper takes the multi-core image processing parallel design model applied to the geological exploration, and designs and implements the multi-core parallel image mosaic algorithm which is based on the proposed parallel design model. This algorithm can take full advantage of parallel computing resources of multi-core processor and shorten the running time of mosaic of large-scale borehole wall expanded images.The main contents of this paper can be summarized as follows:①Discusses research background and significance of the subject, and summarizes the main research topics of multi-core technology at home and abroad as well as describes the concept and state of art of multi-core image processing. Especially, introduces multi-core parallel techniques from different aspects. i.e. the characteristic of multi-core architecture, classification and programming model of multi-core parallel computing, performance evaluation of multi-core parallel processing algorithm, etc.②For the existing methods of parallel design are focused on the specific algorithms such as edge detection and image projection, which are lack of universality, the paper proposes a multi-core image processing parallel design model. Then the paper takes the parallel design of image Fourier transforms algorithm as an example to testify the effectiveness of this model in single-core, double-core, quad-core and eight-core computing environment. The experiment indicates that the proposed multi-core parallel design model has a good scalability, and this model can extend the space of application in image processing.③For mosaic algorithm of large-scale borehole wall expanded images is very time-consuming, the paper takes the parallel design model applied in practical engineering. According to the influence on parallel performance owing to partition of parallel granularity, the paper completes parallel modeling of mosaic of borehole wall expanded images on two kinds of granularity including photo-level and task-level. Finally, the paper implements task-level multi-core parallel image mosaic algorithm of the borehole wall expanded images in the multi-core computing environment. The performance analysis on it shows that the algorithm has good speedup and efficiency on mosaic of large-scale borehole wall expanded images, and improves performance significantly.④The paper has realized the borehole wall multi-core image mosaic system based on the need of actual geological exploration.Practical application on geological exploration shows that the parallel model proposed in this paper can be applied elastically, be popularized expediently, and has versatility. At the same time, there are not only important practical value, but also high theoretical significance in the fields of image processing. |
PDF Full Text Request