JPS62160593A - Character reader - Google Patents

Abstract

PURPOSE:To improve identifying accuracy by detecting a character area from the picture information, comparing the number of picture elements of high level with that of picture elements of low level finding the former and the latter from the picture information at a detected character area, and outputting them as an identifying parameter adding on a recognized result. CONSTITUTION:The picture information stored on a multivalue picture memory 4 is inputted to a spatial differentiation circuit 5. The spatial differentiation circuit 5 generates differential patterns of the above-mentioned picture information in a lateral and a longitudinal directions. The differential pattern is binary pattern in which patterns of positive polarity and negative polarity generated as multivalue differential patterns are binary-coded with an optional absolute threshold value. The height of a character string 1a is detected from a projecting quantity in the lateral direction, and the character width of the character string 1a is detected from the projecting quantity in the longitudinal direction. And based on an obtained area height and area width, an area detection circuit 7 extracts the character area. In this way, the detected picture information at the character area is inputted to a histogram generation circuit 8. The histogram generation circuit 8 generates the histogram for the number of picture elements from the inputted picture information.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a character reading device that identifies characters of an arbitrary recognition target.

[Technical background of the invention and its problems] A character reading device for reading characters engraved on products on a production line and various characters that exist in everyday life reads characters on a predetermined form. General-purpose character reader (○
Unlike OR), it is necessary to find characters from various miscellaneous backgrounds and perform character recognition.

On the other hand, when identifying such an H1 target, it is important not only to have information about the recognized character code, but also to understand which density level is higher when comparing the character and its background. many.

Taking the reading of a car license plate as a familiar example, information on the shading level of the background and letters is an important factor in distinguishing between a private car with green letters on a white background and a commercial car with white letters on a green background.

[Objective of the Invention] The present invention has been made based on the above points, and its purpose is to improve the identification accuracy of recognition targets by adding shading information of characters and their backgrounds as identification parameters. The objective is to provide a character reading device that can contribute to the

[Summary of the invention]

The present invention detects a character area from image information obtained by an observation means, recognizes a character from the image information of the detected character area, and also detects a high-level pixel count and a low-level pixel count from the image information of the detected character area. The present invention is characterized in that the level pixel number is determined, the magnitude of both pixel numbers is compared, and the result of this comparison is output as an identification parameter in addition to the above-mentioned recognition result.

〔Effect of the invention〕

According to the invention! When identifying a recognition target, not only the character recognition result but also the shading information between the character and the background can be used as identification information, so it is possible to improve the identification accuracy.

(Embodiment of the Invention) An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows the configuration of a character recognition device according to this embodiment.

A vehicle 1 to be recognized has a character string 1a, which is a vehicle registration number, stamped on its license plate. The television camera 2 scans the vehicle 1 and reads the character string 1a. A signal from the television camera 2 is A/D converted by an A/D conversion circuit 3. This is stored in the multivalued image memory 4 as image information. The image information stored in the multilevel image memory 4 is input to the spatial differentiation circuit 5.

The spatial differentiation circuit 5 creates horizontal and vertical differentiation patterns of the image information. This differential pattern is a binary pattern obtained by binarizing a pattern of royal authority and negative polarity generated as a multi-value differential pattern using an arbitrary absolute threshold value. The projection circuit 6 calculates the amount of projection in each direction for the differential pattern in each of these directions. The area detection circuit 7 quantizes the obtained projection amount using a predetermined threshold. and,
The height of the character string 1a is detected from the horizontal projection amount, and the character width of the character string 1a is detected from the vertical projection amount.

Then, the area detection circuit 7 extracts a character area based on the obtained area height and area width.

The image information of the character area detected in this way is input to the histogram creation circuit 8. The histogram creation circuit 8 creates a histogram of the number of pixels relative to the density value from the input image information.

In other words, if we compare the number of pixels that make up a character in a character area with the number of pixels that make up the background, and if we apply the general condition that the number of pixels in the background is greater than the number of pixels in the background, then by looking at the pattern of this histogram, we can You can see which reflectance is higher, level or character level. For example, in the example in Figure 2, the number of pixels on the high density (high reflectance) side is
This example shows a case where the number of pixels is greater than the number of pixels on the low reflectance side, and the background is brighter than the characters, as shown in FIG.

Also, Figure 2 shows an example where the number of pixels on the low density side is greater than the number of pixels on the high density side, and the background is brighter than the characters as shown in Figure (a). The histogram is used for calculating the threshold value in the threshold value judgment circuit 9. In the threshold calculation, the valley part of the histogram that is considered to be the boundary between the background part and the character part is detected, and this point is set as the threshold value. The modal method assumes that the histogram has bimodal characteristics (in this case, a character line segment and a character background area vi, >), and that two local maxima separated by a certain distance The density value that is between the two maximum values and takes the minimum value is set as the optimal threshold value 2.As a result, in FIG.
In FIG. 4B, threshold 2 is set for the higher density. Once the threshold value 2 is determined, the number of pixels with a higher level than the threshold value 2 is stored in the high level pixel number storage circuit 10, and the number of pixels with the same lower level is stored in the low level pixel number storage circuit 11. The pixel numbers stored in both pixel number storage circuits 10 and 11 are compared by a comparison circuit 12. When the number of high-level pixels exceeds the number of low-level pixels, the comparison circuit 12
1°' is output, and in the opposite case, 11011 is output. Therefore, for example, in FIG. 2(a), the number of high-level pixels exceeds the number of low-level pixels, so the output of the comparison circuit 12 becomes 1'', indicating that the vehicle 1 is a private car. In (a), the number of high-level pixels is less than the number of low-level pixels, so the output of the comparison circuit 12 becomes O'',
It can be seen that vehicle 1 is a commercial vehicle.

On the other hand, the multivalued image information of the character area is also provided to the binarization circuit 13. The 2m conversion circuit 13 converts the multivalued image information into two values using the threshold value calculated by the threshold value determination circuit 9. The binarized image information of the character area and the output from the comparison circuit 12 described above are input to an exclusive OR circuit 14. The exclusive OR circuit 14 inverts the output from the binarization circuit 13 only when the comparison result is O''.

Therefore, for example, when the reflectance of the background is higher than that of the characters, the number of high-level pixels exceeds the number of low-level pixels, so the output of the comparator circuit 12 becomes "1", and the binarized output is output as is. On the other hand, when the reflectance of the background is lower than that of the characters, the number of high-level pixels is less than the number of low-level pixels, so the output of the comparison circuit 12 becomes "0".
The binarized output is inverted and output. The recognition circuit 15 is
The input character image is compared with a standard pattern prepared in advance, and the recognition result of the character string within the area is output to an external device (not shown). On the other hand, the output from the comparison circuit 12 described above is also output to an external device as a pair of data with the recognition result. Therefore, the external device can use not only the recognition character code but also the m-light information as information for identification, and in this example, it is possible to distinguish between commercial cars and private cars. be able to.

In addition, in the above embodiment, only the detection of the shading ratio of the background and the character line segment was explained, but when a human being directly checks the recognition result, the obtained shading information can be used to display the recognition result in color. You can also do it.

Further, although black and white shading information is used in the above embodiment, it is also possible to further improve identification accuracy by using a color two-dimensional sensor.

Furthermore, in the above embodiment, detection was performed using the area of the character line segment and the background area, but the same effect can be obtained by comparing a part of the character line segment and a part of the background area.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a character recognition device according to an embodiment of the present invention, and FIGS. 2 and 3 are diagrams showing objects recognized by the device and corresponding histograms obtained by the device. It is. 1... Vehicle, 2... Television camera. Applicant's agent Patent attorney Takehiko Suzue (a) (b) Figure 2 1 ,,l fake (a) (1)) Figure 3

Claims (2)

[Claims] (1) a character detection means for detecting a character area from image information obtained by the observation means; a character recognition means for recognizing a character from the image information of the character area detected by the character detection means; a comparison means for determining the number of high-level pixels and the number of low-level pixels from the image information of the detected character area and comparing the magnitude of both numbers of pixels, and pairing the comparison result of this comparison means with the recognition result of the recognition means. 1. A character reading device comprising: means for outputting data as data. (2) A patent characterized in that the comparison means is equipped with a histogram creation means for creating a histogram of the number of pixels with respect to density from image information of a character area, and means for determining a density threshold from the histogram. A character reading device according to claim 1.