Geometric hashing (GH) is a general model-based recognition scheme. GH is widely used in the industrial products assembly and inspection tasks. The aim of this study is to speed up the geometric hashing pattern recognition method for the purpose of real-time object detection applications. In our method, a pattern is decomposed into some sub-patterns to reduce the data number in hash table bins. In addition, the sub-patterns are recorded in a plurality of hash tables. Finally we improve the recognition performance by combining with image pyramid and edge direction information. To confirm the validity of our proposed method, we make a complexity analysis, and apply our method to some images. Both complexity analysis and experiment evaluations have demonstrated the efficiency of this technique.

The aim of this study is to improve the visual quality of x-ray CR images displayed at general displays. Firstly, we investigate a series of wavelet-based denoising methods for removing quantum noise remains in the original images. The denoised image is obtained by using the scheme of wavelet thresholding, where the best suitable wavelet and level are chosen based on theory analysis. Secondly, the image contrast is enhanced using Gamma correction. Thirdly, we improve unsharp masking method for enhancing some useful information and restraining other information selectively. Fourthly, we fuse the denoised image with the enhanced image. Fively, the used display is calibrated, so that it could offer full compliance with the Grayscale Standard Display Function (GSDF) defined in Digital Imaging and Communications in Medicine (DICOM) Part 14. Finally, we decide parameters of the image fusion, resulting in the diagnosis image. A number of experiments are performed over some x-ray CR images by using the proposed method. Experimental results show that this method can effectively reduce the quantum noise while enhancing the subtle details; the visual quality of X-ray CR images is highly improved.