Canon Kabushiki Kaisha

Tetsuomi Tanaka

Shugoro Ueno

Hiroaki Ikeda

Osamu Yamada

Toshiaki Yagasaki

Makoto Takaoka

Yuji Suzuki

Taeko Yamazaki

Yukari Toda

Ichiro Umeda

Shigeo Fukuoka

Masami Kugai

Masayuki Kiriyama

Image compression that keeps a balance between improvement in visual recognizability by highlighting a character color of a reversed character and a reduction in an apparent difference between an input image and a restored image is implemented. An image processing apparatus including a specifying unit configured to specify a character within an image, a setting unit configured to set a representative color to the specified character by using colors of pixels constituting the specified character, a determination unit configured to determine whether the specified character is a reversed character, and a correction unit configured to correct the set representative color in the case where the specified character is determined to be a reversed character.

Image processing apparatus, image processing method, and storage medium

An electronic document producing device has a correcting unit for correcting distortion of a first image to obtain a correction image, and a character recognition unit for executing character recognition processing on a plurality of character images contained in the correction image to obtain text data. The device also has a unit for finding a base line of each character row in the first image, and a unit for finding a relative position from the base line in regard to each character image in the first image. The device also includes a producing unit for producing an electronic document including the text data and the first image, wherein a position of the text data is described based on the relative position from the base line.

Electronic document producing device, electronic document producing method and storage medium

An image processing apparatus according to the present invention has a determining unit for determining a size of an image, a resolution-setting unit for setting a resolution based on the size of the image, and a recognizing unit for recognizing whether text is included in the image having the determined resolution. Therefore a text portion in the image can be recognized efficiently.

Image processor, image processing method and storage medium

An apparatus for document conversion that are capable of facilitating conversion of document image data to an electronic document having table of contents data even with a limited storage resource. The document image analysis sectionextracts character regions from a document image. The contents/index/footer conversion sectiongenerates table of contents data based on the extracted character regions and page numbers of the character regions. An electronic document having a table of contents is generated based on the document imageand the generated table of contents data. Link information is added to respective ones of items in the generated table of contents data for linking the items in the generated table of contents data with corresponding positions in the electronic document in which the items are described.

Apparatus, method and system for document conversion, apparatuses for document processing and information processing, and storage media that store programs for realizing the apparatuses

A multivalued original image is converted to a high-resolution image by interpolation processing and the resultant high-resolution image is subjected to binarization processing to obtain a high-resolution binarized image. This is followed by extraction of a plurality of text regions for every text color, as well as position information and text color information of each text region. First compressed data of the text regions is generated by applying compression processing to the high-resolution binarized images at the positions corresponding to the text regions extracted. Second compressed data is generated by filling text regions in the original image with a prescribed pixel value and applying compression processing to the image obtained by such filling. Compressed image data of the original image is then generated, this data including the first compressed data and the second compressed data as well as the position information and color information of each text region.

Information processing apparatus, method of controlling same and computer program

An image processing apparatus generates binary image data by performing binarization processing on input image data and generates edge image data by performing edge extraction processing to the input image data. The image processing apparatus acquires a first area identification result by performing area identification processing on the generated binary image data and acquires a second area identification result by performing area identification processing on the generated edge image data. In addition, the image processing apparatus generates a merged area identification result by merging the first area identification result and the second area identification result and generates an output binary image by performing partial binarization processing on the input image data according to each character area included in the generated merged area identification result.

Image processing apparatus, image processing method, and computer program

An image processing apparatus processes an input image, in which an edge image is created by executing an extracting operation on the input image, and a binary image is created by performing a thresholding operation on the input image. A first partial thresholding unit creates an output binary image by extracting a character area from the input image, which corresponds to a character area identified from the created edge image, and performs a thresholding operation on the extracted character area. A second partial thresholding unit creates an output binary image by extracting a character area from the input image, which corresponds to a character area identified from the created binary image, and performs a thresholding operation on the extracted character area. An output unit outputs the output binary image created by said first partial thresholding unit and the output binary image created by said second partial thresholding unit to a memory, wherein the first partial thresholding unit determines whether a character in the extracted character area is a reversed character, and if it is determined that the character is a reversed character, creates an output binary image by reversing and performing the thresholding operation on the character area.

Image recognition system utilizing an edge image and a binary image

There are provided an image coding method and apparatus and image compression method and apparatus which can properly compress a color document image with a high compression efficiency. First of all, a color reduction processing unit () converts each pixel of a color document image into an index assigned in correspondence with a color value, and generates color information () containing the number of pixels for each indexed color and an index color image (). A color information sort unit () and the like determine, for example, a color value corresponding to an index of the maximum number of pixels as a background color. An order determination unit () determines an order. A binary image generation compression unit () generates a partial binary image for each index and compresses the binary image according to the order. A data integrating unit () generates compressed image () by integrating background data () and the compressed data of the partial binary image.

Image compression method and apparatus, and image coding method and apparatus

This invention has as its object to provide an image processing apparatus and method which can efficiently compress an image while maintaining high image quality, and a storage medium for implementing the method. To accomplish this, an apparatus has a text area detector () for extracting a text area from multi-valued image data, and generating text area coordinate data () as position data of the text area, a binarization unit () for generating binary image data () by binarizing multi-valued image data of the text area, a color computation unit () for generating text color data () by computing a representative color of the text area, a text paint unit () for generating non-text multi-valued image data A by converting multi-valued image data of the text area using multi-valued image data of an area other than text, a reduction unit () for generating reduced non-text multi-valued image data by lowering the resolution of the non-text multi-valued image data, a JPEG compression unit () for compressing the non-text multi-valued image data by JPEG, and an MMR compression unit for compressing the binary image data of the text area by MMR.

Image processing apparatus and method

An image block is detected by raster scanning an input image and chasing surrounding black pixels around a detected black pixel as the start point in sequence. Further, by judging the feature of the detected image block, the attribute of the image block can be judged. Also, by comparing the detected image block with a predetermined window, it is possible to determine whether or not the image block is noise, whereby image processing of higher accuracy can be effected.

Image processing method and apparatus for discriminating an input image area or an attribute of image information in the discriminated area

When an image is to be rotated, the image is divided into blocks each having a predetermined size, and rotation of the image is realized by rotation of the respective blocks and rotation in units of blocks over the entire image. In this case, lines included in each block are moved to a predetermined region so that the lines are continuous in the order of lines. In the region in which the lines are moved as described above, predetermined digit positions in the respective lines appear at intervals of pixels for one line irrespective of the digit positions. After the respective digits are set to have a predetermined positional relationship between lines, since a logic operation is simultaneously performed for a plurality of lines, a plurality of pixels can be simultaneously moved to have a positional relationship after rotation. Also, a plurality of sets of pixels having the positional relationship after rotation can be similarly, simultaneously moved.

Image processing method and apparatus for rotating an image in an area

An image processor determines a binarization threshold value used for binarizing a multi-valued image, and performs binarization processing on the multi-valued image. The multi-valued image is input line-sequentially, the binarization threshold values are determined sequentially in units of predetermined size blocks each including a plurality of pixels, and the multi-valued image data is binarized line-sequentially. The inputting of data, determining the threshold value, and binarization of the multi-value image data is performed by pipeline processing.

Image processing method and apparatus for determining a binarization

An image processing apparatus comprises: two table memories; a producing circuit for executing a process according to a kind of each character image and for producing contents in the table memories; a character memory to store the image as an object to be recognized; a feature extractor for adding the data in the table memories and for executing a character feature extracting process after normalization of the character image by the bit data in the character memory corresponding to the addition result; a memory to store data indicating to which one of binary values of 1 and 0 the character image portion belongs; and a circuit for inverting the image data from the character memory to the feature extractor by the stored data and for outputting. The apparatus receives image data of characters of various type styles and outputting directions and can convert into the normalized image data such as to be easily processed.

Image processing method and apparatus

There is disclosed a character recognition method, and an apparatus therefor, capable of improving the accuracy of character recognition employing character normalization and direction index counting. The original image information is read with a resolving power matching the character size, so that the result of direction index counting is not distorted by the character size change in the normalization.

Growing community of inventors

Kawasaki, Japan

Tetsuomi Tanaka