| The technology of Imax (mean Image maximum) film are especially the inevitable result of continuous development of the people's appreciate. Imax film with its huge screen, realistic three-dimensional scenes and shocking the acoustics conquer an increasingly more and more "nit-pick" the audience.Then,Imax film's popular is still a long way to go. In addition to the high theater building and screening equipment, the most important problem Imax film is its source. Including the special 65mm, 15-film, we must also demand heavy shooting equipment for an Imax film shooting.Therefore, Imax films in general is bound to lead to the film itself will not be too long due to the filming equipment and the film itself. Show time of Imax films is generally only around one hour. Most of the Imax films are technology and scenery, and really shooting the story films are rarely. Imax film into the real audience's attention is due to a specific processing technology named DMR technology (Digital Re-mastering Technology). This technology can handle ordinary 35mm to 70mm film into the Imax film because of Imax DMR film and processing technology invented by Imax Canada company.So, DMR specific technology is in the privacy to the external. In order to developing of China's film industry, it is necessary to break through the restrictions DMR technology and find our own better image processing methods though. Although we do not know the core of the DMR technology, the basic approach is clear, that is, the storage to a specialized computer workstations by the ordinary 35mm film scanning and then followed by interpolation,enlarge, de-noising and enhancing with a final ordinary printing photos technology, we see the formation of the Imax film.After the scanned Imax image is a film image of bmp format. For obvious reasons, this format picture is much larger than ordinary bmp we often see and two films images for the difference between are small, almost the same, but it is a movement inside the continuity, that is, the two images between before and after have the link relation. This is the difference between the ordinary bmp and the Imax film image, but in the processing, we can still use an ordinary digital image to replace Imax bmp.There are many digital image processing methods, the most classic is that the traditional method of Fourier transform. As a new development of image processing technology, wavelet method is a rookie in well time-frequency performance image processing, as the "mathematical magnifier". With the development of new image processing technology, there will become more and more of image processing names and types. But the most basic processing technology is divided into two categories: firstly, the spatial domain approach and the second is transform domain approach. Spatial domain of the digital image is the data itself or smaller sub-image processing, including neighborhood treatment and points treatment. Transform domain approach is the transforms in the image domain, that is, transform the image data through a switch to other space, the transformation of data processing, and then converted back to images of space to get the necessary results. Generally using Fourier method, discrete cosines transform (Discrete Cosine Transform, DCT) and wavelet transform etc. According to objects and the different treatment methods of the spatial domain and frequency domain, we can see that only a simple method of image processing is not complete, we must combine the two different methods based on the characteristics of these two methods, and then we directly enhance the digital image in spatial, and then de-noising on the basis of improving.Wavelet method has very good time-frequency domain characteristics. In dealing with IMA film images, we first decomposed them with their first MALLT algorithm, and then in different scales of decomposition based on decomposition, we de-noise and enhance the images. In this paper, we use the enhancement method of the gradient transform and Lap lace transform method. We general select threshold above 50 in gradient transform and 10 in the Lap lace threshold. There are many De-noising methods of Wavelet-based Digital Image. In addition to the traditional method of de-noising, there is mainly selected de-noising based on values of the hardware and software of different threshold. According with the value of selected threshold, we first processed images with edge detection, and then reconstructed images based on the image processing, so as to achieve a better effect.In image processing, edge detection is an important aspect. The image edge information directly determines the basic content of the image. There are many images constituted of the edge information. We commonly used method of Sobel edge detection and wavelet transform edge detection method. In this paper, we adopted testing edge information method based on a different threshold selection and software and hardware threshold edge information in different performance aspects of dealing with the edge information. It's an ideal edge detection method.In dealing with this Imax film image, first, we decomposed the image of two-dimensional discrete wavelet decomposition, and then selected different threshold on the basis of image decomposition. And last, we enhanced image with the improved Gauss-lace method according to the threshold. The noise image of enlarged increase, we first carry out DTCWT (Dual-tree Complex Wavelet Transform) de-noising including complex wavelet HMT, then reconstructed by Mallat algorithm after decomposition algorithm, and then used stationary wavelet transform to smooth image, then we can obtain a satisfactory effect and complete the de-noised Imax film image.For the results of assessment after dealing with image, in addition to adopting specialized PSNR (peak signal to noise ratio) and the mean square error, we must take into account the visual experience of human eye.In this paper, because of the good analysis capabilities, limited data redundancy as well as fully meet the conditions of the advantages of remodeling with the DTCWT (Dual-tree Complex Wavelet Transform). We combined relationship of the scale integration between DTCWT methods and complex wavelet domain HMM. And to consider the scale, we adopted correlation filter to complex wavelet coefficient in accordance with neighborhood local adaptive method. Practice has proved that this approach in dealing with the enlarged image of the Imax film, we can achieve more satisfactory results. |
PDF Full Text Request