Image compression in real time by vectorial quantification

摘要

The real-time image compression commences by dividing the image into blocks, or vectors, each grouping a predetermined number of pixels. In each block all pixels are coded by their grey level as numeric values forming components of a vector (V). A catalogue of prototype vectors having as many components as the vector to be coded has been stored previously The absolute value of the errors between corresponding components of the vector and the prototype vectors is computed. The sum of these absolute errors represents the rectilinear distance between the vector and the prototype vectors. The smallest distance is determined and the code representing the closest prototype vector is stored or transmitted.