Images are the main source for humans to obtain information from the outside world. Since the image is available, there is also the technology to repair the damaged image. Inpainting image restoration technology is an ancient art. In the European Renaissance, in order to restore the lost or damaged parts of the art work, while maintaining the overall effect of the work, people began to repair the art work, which is mainly to fill in the cracks or grooves in the work, such work is called Inpainting.
Nowadays, image Inpainting repair technology has become a new active research direction in the field of image engineering. Its purpose is to study and solve how to better detect damaged parts on images, and use image repair according to the effective information around the damaged images. The algorithm recovers automatically. Although some powerful digital image processing software, such as Photoshop, can also perform professional special effects processing on digital images and repair damaged images, it requires sophisticated manual processing by experienced technicians and cannot reach the "computer" The purpose of "smart" automatic processing.
Let's first look at an example of image restoration. Figure 1 (b) is a letter symbol similar to "@", part of which is blocked by obstacles. According to the human visual experience, referring to the shape of the obscured object and the surrounding image information, make a guess judgment, that after removing the obstacle, the characters shown in Figure 1 (a) may appear, or may appear as shown in Figure 1 The characters shown in 1 (c), and the repair results of Figure 1 (a) and Figure 1 (c) may be correct. The computer image Inpainting technology simulates this effect from the perspective of human visual psychology. According to the edge information of the obscured object, it extends and diffuses and connects the boundary in a certain direction to fill the obscured part to achieve visual connectivity. Of course, from the figure, we also see that the size of the image's damaged information (the size of the obstacle) also directly affects people's visual connectivity judgment. From a mathematical point of view, this is caused by insufficient information around the damaged image. The phenomenon that the image repair result is not unique is a morbid problem, so the image Inpainting technology also needs to design a repair algorithm through the limited design of certain assumptions based on computer vision theory to solve this problem.
According to the Bayesian (Besian) structural system, we obtain the perfect image from the incomplete and deformed data based on the "best hypothesis" or "optimal guess" made by the simulated human eye recovery image. The "best assumption" is based on the following two important factors:
1. Image data model: how do we get more existing image data information in the original image.
2. Image pre-model: What kind of image model should the original image be intact? For example, if we are repairing a picture of bananas and apples in a fruit bowl, we will have a pre-impression that it should have a smooth shape, filled with yellow and red.
Therefore, the "optimal guess" is to maximize the probability of the latter in the Bayesian probability model, fill in the missing or damaged part of the image according to certain algorithm rules, so that the restored image is close to or achieves the visual effect of the original image.
The development of image repair technology is mainly concentrated in two directions: image repair based on texture structure and image repair based on non-texture structure, as shown in Figure 2.
[next]
For the Inpainting repair technology of texture structure images, such as wood grain and rock pattern, the core idea is to simulate and generate local textures for filling. Because texture is a reflection of the distribution or characteristics of material constituents, its local shape information can express the commonality of the same texture. For a texture map, any two of the small pieces of texture are similar, so to repair the image using the method of texture synthesis, you can pass: (1) histogram statistics of the color information of each pixel within a set range ; (2) Create a weight table and calculate the degree of certainty of the color information of the relevant pixels; (3) Evaluate and select the most matching pixel for image repair; (4) Repeat in this way until the damaged image is filled, as shown in Figure 3. Among the various algorithms for implementing texture synthesis, we should also extend the scope of research from general textures to directional textures, and further to surface texture synthesis, video texture synthesis and other aspects.
For the non-texture structure image Inpainting repair technology, researchers currently use repair algorithms based on higher-order partial differential equation (PDE) models. The main idea is to use the edge information of the region to be repaired to determine the diffusion information and the diffusion direction, and diffuse from the boundary of the region to the boundary anisotropically. The algorithm can simultaneously fill multiple areas with different structures and backgrounds, and there is no limit to the topological relationship of the areas to be repaired. In addition, Manou et al. Proposed a fast image Inpainting technique based on the above ideas. By determining the direction of the isobaric line at the boundary of the area to be repaired, the corresponding isobaric line is connected with a straight line. To spread within the range of the intensity line to fill the area to be repaired. This method has a good repair effect on simple structure images, and the repair time is greatly shortened.
Image Inpainting repair technology has a wide range of applications, first of all for the repair of cracks in static images and the removal of obstacles. In the printing industry, different types of digital image manuscripts that need to be processed before printing, such as damaged old photos, scratched transmission negatives, or removing obstacles in the image that are not related to the image content, can be simply through Inpainting technology. repair. The user only needs to simply select the repair range, and the computer will automatically fill in the scratches, blank areas, or fill in the background pattern after removing the obstacles, which can greatly reduce the time and manual workload of pre-press image processing, and Further image processing can be performed according to the result of the repair, and the repair effect is shown in FIG. 4.
The second is dynamic image repair. Image Inpainting repair technology can be applied to the film and television industry. For example, for film and television copying, if one of the continuous frames has scratches or stains that need to be repaired, we convert the continuous frames of the film into a digital image sequence, and then according to its Compare the information of the front and back frames, extract useful information from adjacent frames for copy repair. In addition, we can also remove the subtitles in different languages ​​included in the movie through Inpainting repair technology, as shown in Figure 5.
Thirdly, in some fields of photography, Inpainting technology can also be used to eliminate the red-eye phenomenon in digital photos, or when the digital photos taken in stages are stitched together in panoramic images, Inpainting technology can be used to repair overlapping parts in adjacent photos.
Inpainting technology has been concerned by many domestic scholars, and has become an active research field abroad. In short, the technical advantages of Inpainting will inevitably involve more image application fields, such as biomedicine, remote sensing detection, and promote the development of this technology in more scientific fields.
Candle packaging boxes are designed for those handmade, scented candles. Most candle boxes will have an attractive look and can give protection to your handmade or ready-made candles. Cailang Printing Products Co.,Ltd are capable of providing custom made candle packing boxes for candle industry with full colors, sizes and styles. Spot UV , embossing , debossing , hot stamping , glitter finishing . We`ll make your candle packaging look amazing .
If you are still looking for a long-term partner for your packaging , do not hesitate to contact us anytime .
Custom Candle Packaging Box,Luxury Custom Candle Packaging Box,Candle Packaging Box,Rigid Candle Packaging Box
Huizhou Cailang Printing Products Co.,Ltd. , https://www.paperboxs.nl