9 results about "Texture transfer" patented technology

Mesh simplification reduces the complexity of 3D objects while maintaining visual fidelity. While there are effective techniques for mesh simplification for simple meshes, in actual use cases it may be difficult to simplify more complicated meshes. A technical solution retrieves an initial mesh (e.g. a triangle mesh) and constructs an edge list including physical and virtual edges for the initial mesh. The initial mesh is then decimated by iteratively collapsing the edges, prioritized based on a modified quadric error metric that may combine a memory edge quadric error and a memoryless area quadric error. The technical solution may maintain a correspondence map during the decimating and use it to perform texture transfer after simplification based on the correspondence map.

The present invention discloses an AI generation method for visual product appearance, which provides auxiliary reference and inspiration for industrial product appearance designers. When an original product, a target object and an artistic style picture are input, the method deforms the texture appearance of the original product into the shape of the target object, thereby forming a new product, and at the same time transfers the artistic style to the new product, ultimately forming a wonderful visual product appearance. The method is named industrial style transfer, which consists of two parts: a large-scale geometric deformation module and an interest-preserving texture transfer module. Among them, the large-scale geometric deformation module aims to learn the conversion between the shape masks of the original product and the target object in an unsupervised manner to achieve large-scale geometric deformation. The interest-preserving texture transfer module introduces an interest regularization term on the traditional texture style transfer, so that important content details in the original product can be retained as much as possible after the texture transfer.

The invention relates to an image style migration method and device based on similarity measurement guide area self-adaption, which can generate a naturally coordinated stylized image, realize self-adaptive style migration based on a semantic area, enhance the detail capture capability of a model, and generate a visually satisfactory high-quality image while keeping the complete structure of the image. The method comprises the following steps: inputting features extracted from content and style images into a texture migration module guided by similarity measurement, and guiding a Transform decoder to fuse the content and style features by a cross attention module based on a Gaussian kernel function; the method comprises the following steps: respectively dividing a content image Ic and a style image Is into local image blocks, expanding each image block into a feature vector, respectively constructing a content feature matrix Vc and a style feature matrix Vs, sequentially calculating a similarity matrix # imgabs0 # # imgabs1 # between all content blocks and style blocks by using a Gaussian kernel function as an attention mask to modulate cross attention calculation, attention distribution is guided through semantic similarity, and accurate matching of the content and the style area is achieved.

The present invention discloses an embossing texture transfer device, specifically related to the field of embossing devices, which includes a machine body, an embossing chamber provided on the machine body, a pressure-bearing mold connected in the embossing chamber, and an embossing mold structure located above the pressure-bearing mold in the embossing chamber; it also includes an anti-deviation mechanism for limiting the transmission of the printed matter to be printed. The present invention can precisely adjust the distance between the first sliding seat and the second sliding seat, as well as between the first adjusting clamp and the second adjusting clamp, and synchronously adjust the blocking rod and the anti-deviation guiding and pressing wheel, so as to be applicable to the limiting and guiding of printed matter of different widths; after the adjustment is completed, the printed matter is introduced from the incoming and outgoing slots for the printed matter to be printed on one side of the embossing chamber. At this time, the printed matter is limited under the action of the blocking rods on both sides, and at the same time, the edges of the printed matter are pressed under the action of the anti-deviation guiding and pressing wheel, which can prevent deviation during embossing, effectively reducing the rejection rate of the printed matter, greatly improving the printing accuracy and printing effect, and improving the printing efficiency.

The application provides an image generation method based on style content adaptive normalized pose guidance, belonging to the technical field of image synthesis, inputting a person image, selecting a source image and a target image in the person image, generating a target person image consistent with the source image style and the same as the target image pose, first, the pose information transfer is performed in advance by using an aligned multi-scale content transfer network to predict a target edge map, which not only retains the texture content but also can relieve the spatial dislocation. Secondly, the source style feature is gradually transferred to the target pose and reasonable arrangement is realized by using a style texture transfer network, which is realized by a style adaptive normalization generator, the source style feature, the target pose and the edge are mapped in the same hidden space, the consistency of the style texture and the content is enhanced by adaptively adjusting the source style and the target pose, so that the source style feature is retained to enhance the texture details of the target generation.

Approaches presented herein can utilize a network that learns to embed three-dimensional (3D) coordinates on a surface of one or more 3D shapes into an aligned two-dimensional (2D) texture space, where corresponding parts of different 3D shapes can be mapped to the same location in a texture image. Alignment can be performed using a texture alignment module that generates a set of basis images for synthesizing textures. A trained network can generate a basis shared by all shape textures, and can predict input-specific coefficients to construct the output texture for each shape as a linear combination of the basis images, then deform the texture to match the pose of the input. Such an approach can ensure alignment of textures, even in situations with at least somewhat limited network capacity. To unwrap shapes of complex structure or topology, a masking network can be utilized that cuts the shape into multiple pieces to reduce the distortion in the 2D mapping.

A method includes determining a symmetry plane that the divides a mesh into a first side and a second side opposite the first side; performing, on the first side of the mesh, a decimation process to generate a decimated first side of the mesh; performing UV reparameterization on the decimated first side of the mesh to generate a first side UV reparameterized mesh; performing texture transfer on the first side UV reparameterized mesh to generate a first side texture image; determining whether the first side texture image exhibits a degree of texture symmetry with the second side of the mesh that satisfies a condition; and based on a determination that the condition is not satisfied, performing a symmetry texture transfer process between a first texture coordinate on the first side and a second texture coordinate on the second side.

Systems and methods are provided for transferring color from a candidate garment not suitable for a virtual try-on experience for a user to a target garment of an image suitable for the virtual try-on experience. The source and target images for transferring color may be determined utilizing an amount of overlap of 3D segmentation masks for a candidate garment depicted in a candidate source image against the target garment depicted in a target image to identify that the candidate garment is suited for color transfer based on the comparison. Color or texture may then be transferred to the target garment based on image data of the identified candidate garment.

A manufacturing process for seamless grafting of multiple textures on a mobile phone back cover is made through the following steps: Step 1, prepare raw materials; Step 2, transfer printing; Step 3, first texture transfer printing; Step 4, second texture transfer printing; Step 5, perform a release surface treatment on the side with the texture area of the double-texture carrier board in Step 4 to obtain a replication master for double-texture effect replication. The manufacturing process for seamless grafting of multiple textures on a mobile phone back cover of the present invention realizes the grafting of two different textures through the method of exposure and development. The appearance effect is the same as that of photolithography, the transition area is very flat, the splicing of the entire texture is not affected by the position and shape, and it is flexibly applied, ensuring the surface flatness of the mobile phone back cover in the later stage. Multiple textures can be seamlessly grafted together. The entire grafting cost is low and the cycle is short, which greatly facilitates the selection of models for early development.