An image processing apparatus includes an encoding manner storing section that stores encoding manners in association with quantities of characteristics of objects, a characteristic region detecting section that detects a plurality of characteristic regions from an image, and a compressing section that compresses the images of the plurality of characteristic regions by encoding manners stored in the encoding manner storing section in association with the quantities of characteristics of objects included in the plurality of characteristic regions respectively

Image processing apparatus, image processing method and computer readable medium

Off-line Handwritten Character Recognition Using Features Extracted from Binarization Technique☆

Offline cursive script recognition and their associated issues are still fresh despite of last few decades' research. This paper presents an annotated comparison of proposed and recently published preprocessing techniques with reported work in the offline cursive script recognition. Normally, in the offline script analysis, the input is a paper image or a word or a digit and the desired output is ASCII text. This task involves several preprocessing steps, and some of them are quite hard such as line removal from text, skew removal, reference line detection (lower/upper baselines), slant removal, scaling, noise elimination, contour smoothing and skeleton. Moreover, subsequent stage of segmentation (if any) and recognition is also highly dependent on these preprocessing techniques. This paper presents an analysis and annotated comparison of latest preprocessing techniques proposed by authors with those reported in the literature on IAM/CEDAR benchmark databases. Finally, future work and persist problems are highlighted.

Annotated comparisons of proposed preprocessing techniques for script recognition

This paper presents detailed review in the field of off-line cursive script recognition. Various methods are analyzed that have been proposed to realize the core of script recognition in a word recognition system. These methods are discussed in view of the two most important properties of such systems: size and nature of the lexicon involved and whether or not a segmentation stage is present. Script recognition techniques are classified into three categories: firstly, segmentation-free methods or holistic approaches, that compare a sequence of observations derived from whole word image with similar references of words in the small lexicon. Secondly, segmentation-based methods, that look for the best match between consecutive sequences of primitive segments and letters of a possible word similar to human-like reading technique, in which secure features found all over the word are used to boot-strap a few candidates for a final evaluation phase; thirdly, hybrid approaches. Additionally, different feature extraction techniques are elaborated in conjunction with the classification process. In this scenario, implications of single and multiple classifiers are also observed. Finally, remaining problems are highlighted along with possible suggestion and strategies to solve them.

Off-line cursive script recognition: current advances, comparisons and remaining problems

A Mailport that solves a variety of problems caused by the proliferation of Automated Parcel Carriers, including unmanned aerial vehicles (“UAVs”). The preferred embodiment of the invention features an access aperture designed to retrieve and transfer packages to and from Automated Parcel Carriers. The preferred embodiment of the invention also features a storage area for parcels, and a secured aperture to enable intended recipients to receive or send Parcels as appropriate. Various embodiments of the invention may stand alone or otherwise integrate within the structure of a building.

Mailport for automated parcel carriers

Word recognition algorithms are classified into two major groups One is an "analytic" approach of recognizing individual characters, while the other is a "holistic" approach dealing with an entire word image In the former approach, matching scores used to be calculated using heuristic functions, such as an average of confidence values on character recognition In some non-heuristic studies, a stochastic evaluation function is employed, which is a ratio between an "a posteriori" probability and an "a priori" probability ("a posteriori" probability ratio) In this research, a new evaluation function is proposed, which is an improvement of "a posteriori" probability ratio A result of an experiment using real images shows 91% improvement on handwritten word recognition

An Analytic Word Recognition Algorithm Using a Posteriori Probability

A search processing section searches information stored in an address database, using a first character series, e.g., a postal code input through an input device as a search key, for a second character series corresponding to an address. A character recognition processing section performs character recognition with respect to a predetermined area in the image, using a character dictionary, and generates candidates for a character series including a name or designation, a postal code, an address, etc. A character image selection processing section selects a character series corresponding to the searched second character series from the generated candidates. A character image storage section stores correlation between each of the characters constituting the selected character series and a character image thereof. A character dictionary learning processing section performs a learning process with respect to the character dictionary, based on the stored correlation between each character and the character image thereof.

Information processing apparatus having learning function for character dictionary

A pattern recognition method is applied to processing of causing an information processing apparatus to recognize a pattern in a plurality of steps. The information processing apparatus detects recognition candidates which can be recognition candidates of each step. The information processing apparatus expands the recognition candidates of the next step belonging to each detected recognition candidate of each step. The information processing apparatus calculates the evaluation value of each expanded recognition candidate based on an a posteriori probability on condition of all recognition processing results for a recognition candidate which has undergone recognition processing. The information processing apparatus selects recognition candidates based on the calculated evaluation value of each recognition candidate. The information processing apparatus determines a recognition result based on the selected recognition candidates.

Pattern recognition method, and storage medium which stores pattern recognition program

An addressee recognizing apparatus recognizes the addressee of matter to be delivered. A database stores a description format including at least a described position, the number of character lines, and the length of each character line in a sender area on the matter to be delivered. When a scanner obtains an image, a plurality of addressee candidate areas are obtained from the image. The apparatus determines whether the description format for each of the extracted candidates matches the description format for the sender area stored in the database. If the description format for the candidate is determined to match the description format for the sender area, the candidate is prohibited from being recognized as the addressee area.

Addressee recognizing apparatus

In an image compression device, an image input section performs image input processing to input an image having characteristics which are recognition objects, a conversion processing section performs conversion processing to convert a pixel value of each pixel forming the image into a pixel value suitable for coding while retaining the characteristics which are the recognition objects in the image input by the image input processing, and a coding section performs coding processing as compression processing with respect to the image constituted of each pixel converted by the conversion processing.

Takuma Akagi

Papers

An Analytic Word Recognition Algorithm Using a Posteriori Probability

Information processing apparatus having learning function for character dictionary

Pattern recognition method, and storage medium which stores pattern recognition program

Addressee recognizing apparatus

Image compression method and image compression device