35 results about "Portable Network Graphics" patented technology

The present invention uses XHTML and XML to output a varied resolution PNG (Portable Network Graphics) representation of a document. The resulting image file is too large to be displayed on a standard computer monitor without forcing the user to scroll the document horizontally and / or vertically. The present invention manually rasterizes the image to an acceptable resolution for viewing on a standard computer monitor and allows the user to electronically sign the presented SMART Doc.

The invention discloses a method for converting CAD (Computer Aided Design) into parquet data which can be freely combined. The method comprises the following steps: 1, removing noisy points of a pattern according to RGB (Red, Green and Blue) value of each point of a PNG (Portable Network Graphic) pattern to obtain a whole colored pattern without noisy points; 2, based on the processed pattern, and according to the color of the pattern, generating a special data unit for each color; 3, matching special data for each data unit to obtain analytical data in the form of JSON (JavaScript Object Notation) format; and 4, combining data in the form of JSON format and different data units to analyze and construct novel parquet. Through adopting the method provided by the invention, real parquet can be in datamation, so that professional designers can quickly generate desired parquet patterns according to these data and non-professionals can quickly make parquet patterns which can be previously made by designers.

The invention discloses a PNG (Portable Network Graphic) image compression method which comprises the following steps of (1) reading an original image, obtaining RGB (Red Green blue) color space of the original image and converting into HSV (Hue Saturation Value) color space; (2) utilizing image pixel data which are obtained in step (1) to set a palette through an octree method; (3) utilizing the palette which is obtained in step (2) to establish pixel palette index on the image which is obtained in the step (1) to obtain an index image; (4) utilizing the palette which is obtained in the step (2) to perform data compression on pixel data in the index image which is obtained in step (3); and (5) generating a final compressed PNG format image. The PNG image compression method has the advantages of obtaining a perfect PNG format image and being good in processing performance and high in compression ratio at the same time due to the facts that a source image is converted into the PNG format image and unnecessary data blocks are deleted.

The invention discloses a watermark embedding and authentication method for PNG electronic invoice image content protection. It is divided into key class, uniform class and non-uniform class; the content of the image blocks of the three components of RGB is encrypted respectively to generate watermark information, and the corresponding watermark embedding position is selected according to the different types of the component image blocks; the watermark is reconstructed by comparing the image blocks and the The consistency of the extracted watermark is used to judge the authenticity of the image block, and the horizontal neighborhood extension detection method is used to further improve the tamper detection performance. While ensuring that the watermark is invisible, the method increases the size of PNG images less; it has high detection accuracy for tampering, low false detection rate, and can effectively resist tampering such as replacement, addition, deletion, and collage.

The invention provides a picture processing method. The picture processing method comprises the following steps: acquiring property information of an original picture, wherein the original picture is a JPG (Joint Picture Group) picture; and when the area of the original picture is not smaller than an area threshold, converting the original picture into a PNG (Portable Network Graphics) picture, compressing the PNG picture, compressing the original picture, comparing the compressed PNG picture with the compressed original picture to obtain a size deviation between the two compressed pictures, selecting the picture obtained after the original picture is compressed when the size deviation is small and a color value of the original picture is greater than a threshold, otherwise selecting the picture obtained after PNG format compression. In the scheme, the picture is processed according to the property information of the picture, and targeted processing is performed according to a plurality of characteristics such as the picture area, a picture size and the color value, so that the size of the picture is reduced, and the quality of the picture is ensured at the same time. The method is particularly suitable for providing a proper picture format for a user in the processes of uploading, sharing and the like of pictures. The problems of overlarge picture sizes, bandwidth waste, poor image quality and the like due to an improper picture format are solved.

The invention discloses a method for generating an animated portable network graphics picture and a device thereof. The method comprises the steps of: when generation of an APNG picture is required, sequencing M PNG pictures which are required for generating the APNG picture according to a displaying sequence from first to last, wherein M is a positive integer; for aiming at an i-th PNG picture i after sequencing (1<i<=M), respectively determining a changed area in the PNG picture i relative to a PNG picture j (j=i-1) before loss compression, and recording the position information of the changed area; performing loss compression on the PNG picture i, shearing the changed area from the PNG picture i after loss compression according to the position information of the changed area in the recorded PNG picture i, and recording the picture data of the sheared changed area; and generating the APNG picture according to the PNG picture at the first position after sequencing. The method and the device provided by the technical solution of the invention can improve a compression effect.

The invention relates to the technical field of biological information, in particular to a method for identifying gene characters on the basis of Torch supervised deep learning. The technical scheme includes that the method comprises steps of firstly, preprocessing gene PNG (portable network graphic) image information data to enable the gene PNG image information data to conform to Tensor data objects for Torch deep learning model training; secondly, building deep learning models by the aid of Torch, training Tensor training data in deep convolutional neural network models to obtain perfect training data identification results and then outputting tag vectors and weight parameters; thirdly, extracting weight parameters of various connections of a certain category of result tags after all training is completed and inversely coding gene image data information to obtain related SNP (single nucleotide polymorphism) loci of characters corresponding to genes and corresponding weight parameters. The method has the advantages that the characters corresponding to the genes can be conveniently and intelligently classified by the aid of the method, genes with unknown characters can be subjected to self-identification, and the like.

The invention discloses a method and device for optimization processing of PNG (portable network graphic) image resources. The method comprises the following steps: respectively extracting palette data of each PNG image and image data of any one PNG image from N PNG image files with the same pattern; and combining the image data and the palette data of the N PNG images to generate a new PNG image group file. By calling the PNG image group file during the development process of J2ME (Java 2 micro edition) terminal applications, the invention can reduce the data redundancy of the PNG image files in this group, thereby reducing the volume of application package and saving storage resources.