JPH04255078A - Image processor - Google Patents

Abstract

PURPOSE:To obtain an image processor capable of automatically correcting a binary reference value at the time of changing the distribution of luminance. CONSTITUTION:A luminance signal in a fine time is converted into a discrete multi-value, operation for sorting the converted value into several ranks is executed for a fixed time, a reference value for binalization is found out based upon the distribution state of the information stored for the fixed time to respective ranks, and an appropriate binary image is formed from an input signal.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus, and more particularly to the configuration of a binarization device of the image processing apparatus.

0002

2. Description of the Related Art Binarization processing in a conventional image processing device generates binarized luminance information from an input luminance signal using a fixed or semi-fixed reference value, and outputs a binarized image. The configuration was as follows.

0003

[Problem to be solved by the invention] However, the brightness distribution of input signals from television cameras etc. is not constant for each image, and it is insufficient to cope with changes in the brightness distribution due to changes in the subject, lighting, etc. there were. Therefore, it was necessary to timely prepare the binarized image obtained from the input signal so that it would be optimal.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to realize a configuration that can automatically correct a binarization reference value in response to a slight change in luminance distribution.

[0005]

[Means for Solving the Problems] The image processing device of the present invention is an image forming device that binarizes a luminance signal using a reference value to generate image information. The luminance signal is characterized in that it includes an arithmetic circuit that obtains a reference value from the obtained information, and binarizes the luminance signal using the obtained reference value.

[0006]

According to the above structure of the present invention, since the binarization reference value is selected according to the luminance distribution of the image, an appropriate binary image can be obtained even if the luminance distribution changes depending on the illumination or the subject.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram showing the implementation of the present invention when applied to an image processing apparatus. In the figure, 1 is an industrial television camera, and 2 is a circuit that converts a luminance signal into a digital value. 3 is a central processing unit for storing digital values in memory, and 4 is a memory. 5 is a circuit that converts the processing result into a binary luminance signal, and 6 is a monitor television that displays the luminance signal.

[0008] Suppose that when the image input from the camera is directly captured on a monitor television, it will look like the image shown in FIG. The image at this time is divided into grid-like minute sections as shown in FIG. 3, and the brightness of each section is determined by digitally converting the luminance signal corresponding to that section. FIG. 4 is a diagram in which the results of digital conversion of each grid are applied to the original grid. 5 in the figure is a graph showing the brightness, or luminance, of these 25 sections using a histogram. The horizontal axis of the graph represents luminance and the vertical axis represents frequency, and the binarization reference value is determined from the distribution of this graph. First, let A be the brightness located on the rightmost side of the graph,
Let B be the brightness with the highest frequency in the graph. A value is obtained by subtracting the brightness difference between A and B from B, and this value is set as brightness C. A, B
, C can be expressed as a formula: "C=B-(A-B)=2
×B-A", and this value of C is taken as the reference luminance. The brightness of each section is compared with this C, and if it is smaller, it is set as 0, and if it is the same or larger, it is set as 1, and the result is expressed again in a grid form as shown in FIG.

[0009]

As described above, according to the present invention, even when the luminance distribution changes as shown in FIGS. 4 to 7, the image shown in FIG. Instead, it is possible to obtain an image in which the subject and background are accurately separated as shown in FIG.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of an image processing device implementing the present invention.

FIG. 2 is a diagram depicting a raw image captured by a camera.

FIG. 3 is a diagram in which FIG. 2 is divided into a grid.

FIG. 4 is a diagram in which the brightness of each divided section is applied to a numerical value within the section.

FIG. 5 is a histogram depicting the luminance distribution in FIG. 4 with luminance on the horizontal axis and frequency on the vertical axis.

[Fig. 6] Substituting 1 or 0 for each partition as a processing result,
A diagram in which sections in which 0 has been substituted are shown with diagonal lines.

FIG. 7 shows the brightness distribution and its histogram that occur when the camera captures the same subject as in FIG. 4 with different brightness.

8 is a diagram comparing the results obtained by processing the image in FIG. 7 using the value C in FIG. 3 and the result obtained by processing the image by newly determining a reference value; FIG.

[Explanation of symbols]

1 TV camera 2 Analog-digital conversion circuit 3 Central processing unit 4 Memory 5 Digital-analog conversion circuit 6 Monitor television

Claims (1)

[Claims] Claim 1: An image processing device that has a process of binarizing a luminance signal, which has a function of storing information obtained by digitally converting the luminance signal into a multivalued signal, and converts the luminance signal into two values from the stored information. An image processing device comprising an arithmetic function for obtaining a reference value for converting into a value, and binarizing a luminance signal based on the obtained reference value.