JPH01196971A - Background density sensing method for read picture - Google Patents

Abstract

PURPOSE:To eliminate accurately the background by detecting the density of background in response to the density histogram of a read picture. CONSTITUTION:A density data of each picture element of a read picture is outputted from an A/D converter 2 and the density data is counted by a counter 3 for each density and its frequency of occurrence is stored in a memory 4 and the density histogram of one original is generated by the repetition of the processing above. Then the threshold value (background density) to eliminate background is set by the density histogram, a comparator 5 and an arithmetic circuit 6. That is, the density maximum value and the density minimum value are obtained, the tentative background density is detected from the density whose frequency of occurrence is maximum and the density maximum value and the background density is detected from the tentative background density and the density minimum value. Thus, the background of the read picture is eliminated accurately.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for detecting the density level of the background of a read original image.

[Background of the invention]

Documents with a background include those in which characters and symbols are written on a monochromatic color document, and those in which characters and symbols are written on a light color or monochrome picture. Conventionally, in order to remove the background from a document having such a background, a certain amount of the background has been uniformly determined from the maximum density value, and density data beyond that level has been processed as background data.

[Problem to be solved by the invention]

However, this uniform processing method lacks reliability when it comes to originals with uneven backgrounds, and the removal of the background may be incomplete.

In view of such circumstances, an object of the present invention is to enable accurate removal of the background of a read image.

[Means to solve the problem]

For this purpose, the background density detection method of the present invention includes a first step of creating a histogram of read image data, a second step of calculating a maximum density value and a minimum density value from the histogram, and a second step of calculating the maximum density value and the minimum density value from the histogram. The procedure consisted of a third step of detecting the temporary base density from the maximum density value, and a fourth step of detecting the base density from the temporary base density and the above-mentioned minimum density value.

〔Example〕

Examples of the present invention will be described below. FIG. 2 is a diagram showing the base density detection processing block. After the original image is read by the CCD 1, it is converted into a digital signal by the A/D converter 2, and shading correction is performed at the same time. As a result, the A/D converter 2 outputs density data for each pixel of the read image. This density data is counted by the counter 3 for each density, and the frequency is stored in the memory 4, and by repeating this process, a density histogram of one document is created. Note that the document reading for creating this histogram can also be performed by capturing only data of several predetermined points by prescanning. Then, using the density histogram created in this manner, the comparator 5, and the arithmetic circuit 6, a darkness value (background density) for background removal is set. Then, through post-processing, all image data with a density level higher than the darkness value is converted to the highest density value (that is, white level), and the background component is removed.

Next, a method for detecting the background density value will be explained with reference to FIG.

(2) First step: As described above, a density histogram as shown in FIG. 1 is created from each pixel of the read image data.

(2) Second order: The maximum density value of the histogram that intersects line a of 10% of the total number of samples in the read image data is DsaX' and the minimum density value is D8°. This makes it possible to remove noise components that are thought to be included in about 10%.

■, Third step: The density difference between the maximum density value D mix and the density maximum value is 8, and the density difference A is 0.
．． The density values that are twice and 0.5 times higher than the density value DP are respectively set as D+ and Dz, and other density values D:l and D4 having the same frequency as those density values D+ and Dz are determined by density shift. Then, a straight line passing through these density values D3 and D4 is drawn, and the density value at the point of the degree O on the straight line is set as D5, and this is set as a temporary base density value.

■, Fourth step: This temporary base density value and the minimum density value D
Let B be the density difference from min, and let the density value Db lower by 0.4B from the temporary base density value D5 be the base density value, that is, the darkness value TH.

As a result, in this example, all data having a density equal to or higher than the background density value Db is regarded as background data and is converted to white.

Note that, as shown by the dotted line in the human duram in FIG. 1, there are cases where the mountain on the left side is not present. In such a case, the formula 0.4B
The processing is performed by changing the value to, for example, 0.2B.

Furthermore, some histograms have shapes that are significantly different from the histogram shown in FIG. 1, as shown in FIG. In such a case, the temporary base density value D is located outside the curve, so the formula 0 for determining the density values D1 and D2 is
．． 2A, 0.58 are changed to 0.1^, 0.4A, etc., respectively, and the dark value density is determined by the same process.

Further, the histogram does not necessarily have a smooth curve as shown in FIGS. 1 and 3, and if noise components are included, the histogram becomes a jagged line. In such a case, it is desirable to average the jagged lines and convert them into smooth curves before performing the above processing.

In addition, in the case of color originals, R (red), G (green), (
B) The background density is individually detected for image data of each color of blue, and the background is removed.

〔Effect of the invention〕

As described above, according to the present invention, since the background density is detected according to the density histogram of the sampled image, the background can be accurately removed.

[Brief explanation of the drawing]

FIG. 1 is a density histogram for explaining the detection of the background density (dark value) in one embodiment of the present invention, FIG. 2 is a block diagram for detecting the background density, and FIG. 3 is a density histogram of another example. . Agent Patent Attorney Tsuneaki Nagao Figure 2 Figure 3 U v5 [b<'+b

Claims (1)

[Claims] (1) The first step is to create a histogram of the read image data, the second step is to find the maximum density value and minimum density value from the histogram, and the provisional base density is calculated from the maximum density value and the maximum density value. A method for detecting the background density of a read image, comprising: a third step of detecting the background density; and a fourth step of detecting the background density from the temporary background density and the minimum density value.