China
Africa
Explore chapters and articles related to this topic
Multimedia Data Compression
Published in Sreeparna Banerjee, Elements of Multimedia, 2019
JPEG is an image compression [1–3] standard developed by the Joint Photographic Experts Group (JPEG) of the ISO/IEC. It is used for the coding and compression of color as well as grayscale images. It is generally a lossy compression standard. JPEG is an effective image compression standard because (1) image data changes slowly across an image, specially an 8 × 8 data block, and therefore there is much interpixel redundancy; (2) it has been observed that humans are not very sensitive to high-frequency data images, and thus redundancy in data can be removed by transform coding; and (3) humans are much more sensitive to brightness (luminance) than color (chrominance). Thus, a chroma subsampling of (4:2:0) can be used. JPEG compression includes the following steps (Figure 7.6): RGB images are converted to YIQ or YUV and color is subsampled.A DCT is performed on 8 × 8 blocks of the image.Quantization is performed on the pixel values.Subsequently, zigzag ordering and run-length encoding is performed.Finally, entropy encoding (Huffman coding) is performed.
Security and Cryptography in Images and Video Using Elliptic Curve Cryptography (ECC)
Published in S. Ramakrishnan, Cryptographic and Information Security, 2018
An image is defined as a collection of values, where value denotes the intensity of light in the form of an array or matrix. This matrix could be two dimensional in the case of black and white images or grayscale images, or it may be a three-dimensional matrix, as in case of a color image (RGB image). There is a little difference between black and white and grayscale images. In black and white image, there are only two colors, either black or white; on the other hand, a grayscale image contains values in between 0 to 255 where 0 is completely black and 255 is completely white. For example, an 8-bit grayscale image can have up to 255, or 28 shades of black and white, (i.e., 8-bit color depth). A true color or RGB image has a 24-bit color depth which is almost 16 million colors, or 224. Images can be stored in JPEG, PNG, TIFF, BMP, etc. formats. All these formats differ in terms of compression technique, output size of the compressed image, and visual depth of the compressed image. In image compression lossless compression means an image can be reconstructed accurately from the compressed image. Formats like PNG, TIFF, and BMP support lossless compression while JPEG is lossy compression. Table 6.1 shows the difference in various image formats.
Creating the Web Site
Published in Tom Hutchison, Paul Allen, Web Marketing for the Music Business, 2013
Graphics are an important part of any web site. There are some general rules to follow when using graphics on a web page. The three most popular formats for using graphics on a web page are JPEG, GIF, and PNG. JPEG is short for Joint Photographic Experts Group. It is good for photographs and supports 16.7 million colors. The compression actually throws out data to create a smaller file. Sharp edges may appear blurred, so JPEG is not recommended for graphics that contain sharp lines or drastic color changes. It also does not support transparency, so if you want the background of your image to be transparent, JPEG is not the format to use. A progressive JPEG file presents a low-quality image at the first moment of download, and then over several passes it improves the quality.
Secured steganographic scheme for highly compressed color image using weighted matrix through DCT
Published in International Journal of Computers and Applications, 2021
Partha Chowdhuri, Biswapati Jana, Debasis Giri
The Joint Photographic Experts Group (JPEG) [1] digital image format is most popularly used image format nowadays. So, selecting a JPEG image for data hiding is useful and important for image authentication and tampered detection etc. JPEG is a compressed image format and uses transform domain principles. A slight modification in the transform domain may cause more distortion to the cover image. Most of the existing steganographic schemes change one coefficient for hiding one bit of data. So, hiding more data causes changing more coefficients that gradually degrade the quality of the cover image which makes it prone to suspicion. So, we have used the weighted matrix to embed more data by changing only one DCT coefficient. This to increases the robustness as well as the imperceptibility of the stego image.
Low-complexity JPEG quantization table requiring only bit-shift operations
Published in The Imaging Science Journal, 2022
JPEG is an international standardized still image compression technique. It is useful to compress both colour or grayscale images. JPEG is a lossy image compression method. The basic encoding process of the baseline JPEG compression is illustrated graphically in Figure 1. The algorithm needs to perform several steps in order to compress an image. First, an N × N input image is subdivided into blocks of typical size of 8 × 8 pixels, which are then transformed from spatial to frequency domain using the two-dimensional discrete cosine transform (2-D DCT) given as follows: where is the DCT matrix called the transformation kernel and is the corresponding transform-domain output matrix. The transformed coefficients are then quantized, where a real values are mapped to an integer values through a rounding operation [4]: where is the rounding function, denotes the division operating element-wise; and is the quantization matrix given by It is clear from (3) that values are floored; and are respectively given by:
Deep Learning Techniques for Banner Image Classification
Published in IETE Journal of Research, 2022
Chandrodoy Pal, Sudhir Deshmukh, Sunita Dhavale, Suresh Kumar
Many images from several news media sites were also included. As our work is to classify banner and non-banner images, to create non-banner images, we have included images that consist of crowd images obtained from malls, banks, fish markets, etc. The images are stored in .jpg format. Table 1 depicts the distribution of the DIAT banner dataset.