Abstract: The improved image compression method and system optimizes a wavelet transform compression technique by further compressing the transformed data by an edge operation before encoding by a conventional technique.

Abstract: A method and system for automatically determining a compression method to be used on a file. A histogram of the file is constructed in order to determine if there are a predetermined number of consecutive byte patterns below a set threshold. If this condition exists, the file is not an image file and compression can be performed using the lossless Lempel-Ziv compression method. If, over the range of the histogram, the file is above the predetermined threshold, the histogram is examined to determine if the average value in the histogram is in a central portion of the histogram. If it is not, the file does not represent an image and again, the Lempel-Ziv compression method is used. If both the byte patterns and the histogram exceed the predetermined threshold and the average of the histogram is in a central portion, the file is determined to contain an image.

Abstract: An image data coding apparatus includes a coder for coding input binary data in a predetermined coding algorithm and for outputting coded data, a conversion circuit for repeatedly performing a converting operation with respect to binary image data supplied from an external unit, a determination circuit for determining a number of times which the converting operation should be repeated in the conversion circuit based on a condition in which, when data obtained by repeating the converting operation the number of times in the conversion circuit is coded by the coder, a number of codes representing coded data output from the coder is minimum; and a circuit, for supplying to the coder, as the input data, converted data obtained by repeating the conversion operation the number of times determined by the determination circuit. The image coding apparatus outputs the coded data and data representing the number of times the converting operation is repeated.

Abstract: An encoding/decoding method for carrying out conversion between run lengths and corresponding run length codes is provided. The run length codes has a format including a first part for storing first information and a second part for storing second information. The first information indicates the length of the code and the second information indicates the corresponding run length. The total number of bits of each of the run length codes is set as an integral multiple of a predetermined number of bits. The present encoding/decoding method is suitable for use in an image processing system, such as a facsimile machine.

Abstract: Correlation between data in two adjacent scan lines is sensed. When the degree of correlation is above a predetermined value the data is compressed by means of two dimensional compression. When the degree of correlation is below the predetermined value the data is compressed by means of one dimensional compression. Typically, both the one and two dimensional compression systems are based on run length encoding. After a predetermined number of scan lines have been compressed by two dimensional compression, one scan line may be compressed by one dimensional compression to prevent accumulation of error.

Abstract: Binary data is generated by a facsimile scanner or the like. A binary run length code indicating a number of successive "1" or "0" bits in the data is formed. A shift code comprising all "0" or "1" bits in a number equal to the number of the run length code is appended to the run length code only when the run length code is longer than a preceding run length code. A shorter run length code is indicated by inserting logical "0" bits in vacant high order places of the run length code so as to make the run length code equal in length to the previous run length code.