SU842870A1 - Device for recognizing symbols - Google Patents

Description

(54) DEVICE FOR RECOGNITION OF SYMBOLS

one

The invention relates to the field of automation and computer technology and can be used to automatically read and recognize sequences of characters having different inter-character spacing.

A device for recognizing I symbols is known, comprising a reading and conversion unit, the output of which is connected to the input of the classification unit, a block of character width matching with a standard interval, a memory unit, one of the outputs of which is connected to the input of the control unit, a classification analysis unit, the unit specifying the interval and block fixing spaces in the line, the output of which is connected to the output of the reading block and. conversion, and the output with one of the inputs of the interval setting unit, the other inputs of which are connected respectively to the output of the character width matching unit with a standard interval, the input of which is connected to another output of the memory unit, and to the output of the classification analysis unit whose control input and the control input of the control unit are connected to the corresponding outputs

the interval setting unit, the inputs of the classification results analysis unit are connected to the outputs of the classification unit, and the input to the memory unit input.

The disadvantage of this device is the lack of accuracy of the classification when reading characters.

The aim of the invention is to increase the reliability of recognition.

This goal is achieved in the device by what is in it. the classification results analysis block contains a buffer memory node, the input of which is connected to the input of the block, and a series-connected threshold node, an adder and a computation node, the outputs of the buffer memory node are connected

S respectively to one of the inputs of the threshold node, the adder, the computing node and the output of the block whose control input is connected to another input of the computing node, as well as the fact that in it the classification block contains a buffer memory node,

0 whose input is connected to the input of the block, comparison nodes, the inputs of which are connected to the outputs of the buffer memory node, and the outputs through serially connected

one of the threshold nodes, coding nodes and another threshold node with block outputs.

The invention is illustrated in the drawings, where in FIG. 1 shows the block diagram of the device; FIG. 2 is a block diagram of the analysis of the results of classification, and FIG. 3 - classification block scheme.

FIG. 1 denotes: a carrier of 1 character sequence, a read 2 and transform block 2, a classification block 3, a block 4 for fixing gaps in a line, a block 5 analyzing the classification results, a memory block 6, an interval specifying block 7, a block 8 for matching character width with a standard interval and control unit 9.

The classification analysis analysis unit 5 (FIG. 2) contains a buffer memory node 10, a threshold node II, an adder 12, and a computation node 13.

The classification unit 3 (Fig. 3) contains a buffer memory node 14, the input of which is connected to the input of the classification unit 3, comparison nodes 15, a group of threshold nodes 16, encoding nodes 17, and an additional threshold node 18.

The device works as follows.

The carrier 1 is moved using the transport mechanism (not shown in Fig. 1) in front of the optical system of the read and convert unit 2 line by line.

Information signals from the output of block 2 are received in block 3 of the classification and block 4 fixing gaps in the line.

The node 14 of the buffer memory of the classification unit 3 accumulates the signals arriving at its input, and in the comparison nodes 15 for each symbol the reference values are determined, which are received through the threshold nodes 16, the encoding nodes 17 and the additional threshold node 18 to the corresponding outputs of the classification unit 3.

In the classification block 3, the most likely character class is determined and the corresponding reference value is fed to the inputs of the classification result analysis block 5, and the coded class signal is fed to the 6jjOKy 6 memory.

From the incoming sequence of reference values, the block 5 of the analysis of the results of classification selects signals for a certain number of cycles corresponding to a different inter-character interval. The node 10 of the buffer memory of block 5 then passes the data entered into it to the threshold node 11, which are then mapped in adder 12 and transmitted to the computational node 13, the control input of which is connected to one of the outputs of the interval setting block 7, which is intended to determine the beginning of the subsequent inter-character interval. The signal corresponding to the duration of this interval limits the time range of the block 5 for analyzing the results of the classification and is generated in the following two ways.

The first method is that the block 4 fixing spaces in the line generates a signal of the white area and thereby determines the beginning and end of the recognition interval. A white area signal is generated when a certain number of readable signals do not contain or contain only insignificant information about the black area.

In this case, the control signals of the classification results analysis unit 5 and the sign width matching unit 8 with the standard interval do not affect this identification interval.

This method of starting and determining the duration of the recognition interval can be applied at the beginning of a string of characters and in the case when there is a white area between two adjacent characters. Thus, the first method serves to synchronize the position of the recognition signals in natural separation points formed by white areas.

The second method is as follows.

The classification analysis analysis unit 5, based on the obtained reference values, generates a certain cycle causing the corresponding identification interval to start, which can be pre-fixed or variable, which is effective when adjacent characters are touching or overlapping.

At the end of the recognition interval, the control signal 9, which is associated with the memory unit 6 and the interval setting unit 7, is transmitted to the external device via the control unit 9, which is connected to the memory unit 6 and the interval setting unit 7.

Compared with the known device, the proposed device for character recognition allows to increase the reliability of character recognition with different inter-character intervals.

Claims (2)

1. A device for recognizing characters according to the author, St. No Sh2915, characterized in that, in order to increase the recognition accuracy, there the classification analysis block contains a buffer memory node, the input of which is connected to the input of the block, and a serially connected threshold node, adder and computation node, outputs of the buffer memory node are connected respectively, to one of the inputs of the threshold node, adder, computational node and the output of the block whose control input is connected to another input of the computational node. 2. A device for recognizing symbols according to claim 1, characterized in that the classification block contains a buffer memory node whose input is connected to the input of the block, comparison nodes whose inputs are connected to the outputs of the buffer memory node and the outputs through one of the threshold nodes connected in series, coding nodes and another threshold node with outputs block. Sources of information taken into account in the examination 1. USSR author's certificate No. 682915, cl. G 06 K 9/00, 04.06.75.