JPH0679331B2 - Image signal binarization processing method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention binarizes an image signal obtained by photographing a work for the purpose of inspecting the function of a display provided on a product (work). Regarding the method.

As a method of automatically inspecting the function of a display such as a liquid crystal display (LCD) equipped in various electronic devices, the display is CC
An image signal is captured by an image pickup device such as a D camera to obtain an image signal in the form of an electrical signal, and the brightness of each pixel in the image signal is compared with a predetermined slice level to perform binarization of ON or OFF. An automatic determination is made whether the display segment is on.

The present invention relates to an image signal binarization processing method therefor.

[Conventional technology]

FIG. 1 is a system for automatically inspecting the display function of a work 10 equipped with a liquid crystal display 12 as an example of a system for automatically inspecting the function of a display, which uses a CCD camera. 1 illustrates a system for photographing and analyzing a portion of 10 indicators 12.

The work 10 is carried by a belt conveyor (not shown), contacts the stopper 14, and stops at a predetermined position. Indicator 12
The CCD camera 16 is installed at a position where the entire image can be captured, and its output is 8-bit analog / digital (A / D)
Is converted into an 8-bit digital signal in the converter 18,
Stored in RAM20. Each address of RAM20 is CCD camera 16
Corresponds to each pixel of the image signal captured. That is, for example, when the number of pixels per 256 scanning lines is 512, it has a capacity of 256 × 512 bytes, and each byte corresponds to one pixel. The personal computer 22 reads the digitized image signal by designating a predetermined address to the RAM 20 in order to obtain the brightness of the pixels in a plurality of predetermined measurement areas on the image, and the readout is performed. By comparing the value of the signal with a predetermined slice level (binarization level), it is binarized to determine whether the display in each pixel is on or not, and to judge the quality of the display function. .

FIG. 2 is a diagram for schematically explaining the binarization processing in the system of FIG.

The shaded portion 100 in the column (a) of FIG. 2 represents the display segment that is turned on, and the window 102 that defines the measurement area is set across the display segment 100. In column (b), the horizontal axis represents the position of each pixel in the window 102, and the vertical axis represents the brightness. The curve plots the brightness of each pixel in the window 102, and Horizontal lines C to G represent five ways of taking slice levels. Of these, E is the most appropriate slice level and can be binarized with the same width as the actual width of the segment 100, as shown in column (e). If the slice level is too high like D, 2 as shown in column (d).
The width of the binarized signal becomes wider, and the number of pixels determined to be “1” becomes larger than the true number of pixels, resulting in an erroneous inspection result. At the level indicated by C, the entire window 102 is determined to be "1". On the contrary, if the value is set lower than an appropriate level like F, the number of pixels which are "1" becomes small as shown in the column (f), and when it becomes G, it disappears at all.

The vertical axis in the column (b) of FIG. 2 is oriented in the direction corresponding to the lightness. However, in software processing to be described conversely, the value representing the brightness has a value in the opposite direction as it becomes brighter. The smaller the value, the larger the value as it gets darker. Since the slice level value is also the same, the larger the slice level value, the smaller the number of pixels, and the smaller the slice level value, the larger the number of pixels.

Liquid crystal displays are generally equipped with one or more light sources to illuminate the display from the side or back of the display to make it easier to see in dark places. The binarization processing is performed on the image obtained by photographing the display device illuminated by the light source.

Since these light sources are generally point light sources, it is difficult to illuminate the entire display with uniform illuminance. Therefore, a standard product illuminated with a light source having a standard brightness is measured in advance and an appropriate slice level is determined for each window as X ₀₁ , X ₀₂ , X ₀₃ . An appropriate slice level is a level as shown in FIG. 2E. As shown in Fig. 2, if the width of a segment when binarized by an arbitrarily defined slice level is different from the true width, the slice level value is raised or lowered so as to approach that value, within a certain range. If they match, the slice level at that time is an appropriate slice level.

When binarizing the image signal of the work to be inspected, the binarization level X ₀₁ , X ₀₂ , X defined for each window
Binarization processing is performed based on ₀₃ ...

Further, since the brightness of the light source varies slightly depending on the product, it is necessary to correct it. This has been conventionally corrected as follows. First, a specific window is defined as a correction window, and the binarization level determined as described above for a work illuminated by a standard light source is stored as X ₀ . The binary level of the correction window is similarly set for the work to be inspected. The value of ΔX which this is calculated by X _'0 Tosureba formula _{△ X = X' 0 -X 0} (1) constant represents the change in the brightness at the location of the correction window. Conventionally, this change
X was applied to other windows as it was to correct the binarization level. That is, the binarization level in other windows uses the value calculated by the formula X ₀ i + Δ (i = 1,2,3 ...) (2).

[Problems to be Solved by the Invention]

The above method gives almost satisfactory results when there is only one light source. However, when there are two or more light sources, there are cases in which they cannot independently be satisfactorily corrected because they have independent variations in brightness.

As an example, consider the case of illuminating from two light sources. When one light source is brighter than the standard and the other light source is dark, the windows in the vicinity of the light source on the bright side are strongly influenced by the light source, and therefore the slice level needs to be lowered. However, when the correction window is affected by both light sources to the same extent, the effects from both light sources are offset and the correction value becomes small, so that the slice level cannot be set appropriately.

Therefore, an object of the present invention is to make it possible to appropriately correct the variation in the brightness of each light source in an image signal obtained by photographing a display device illuminated by two or more light sources and perform binarization.
It is to propose a binarization method.

[Means for Solving the Problems]

The present invention has been made in view of the above objects, and the gist thereof is to define a plurality of measurement regions on an image obtained by photographing a workpiece illuminated by a plurality of light sources, and to measure the number of pixels in each measurement region. In the binarization processing method of an image signal, which binarizes the brightness of the pixel by comparing the brightness and the slice level set for each measurement region, (i) is illuminated with a light source of standard brightness The brightness of each of the plurality of light sources is measured from the image signal of the captured standard work to obtain the measured value of the brightness of each standard light source, and (ii) one of the light sources of standard brightness is measured. The contribution of the brightness from each light source in each of the plurality of measurement regions is measured from the brightness in each of the plurality of measurement regions on the image signal obtained by photographing the standard work illuminated by only the light source, and (iii) About inspection work The brightness of each light source is measured by the same means as the standard brightness measurement of each light source, and (iv) the standard brightness measurement value of each light source and the brightness measurement value of each light source. And a contribution value of the brightness of each light source in each of the measurement areas, a correction value for correcting the fluctuation of the brightness of the light source is calculated for each measurement area, and (v) the slice level is corrected with the correction value. (Vi) The brightness of the pixel in each measurement region is binarized by the corrected slice level.

[Work]

The effect of each light source on each measurement area (window) is measured in advance, and the effect on the correction value is estimated for each light source according to the brightness of each light source to correct the variation in the brightness of multiple light sources. It becomes possible to do.

〔Example〕

FIG. 1 is a block diagram of a display function inspection system to which the method according to the present invention is applied, and FIG. 2 is a diagram for explaining binarization of an image signal. Since both figures have already been described in the section of the description of the prior art, description thereof will be omitted.

FIG. 3 is a flowchart of the preprocessing part for a standard work in the binarization processing method according to the present invention.

After setting the standard work in place (step a),
Figure 5 to represent such two light sources A, B together on, from an image signal captured in this state, the light source A, obtains the brightness Y _A and Y _B of B. The brightness Y _A, processing for obtaining the Y _B is performed in a similar manner as appropriate binarization level set as described above treated.
That is, in FIG. 5, the brightness near the light sources A and B changes as a function of the distance from the center of the light source. Therefore, if the slice level is set to the smaller value side, that is, the brighter side, a substantially circular binarized image centered on the center of the light source can be obtained. Therefore, the slice level values Y _A and Y _B from which the binarized images S _A and S _B having a predetermined area are obtained are the light sources A and
It is a measure to express the brightness of B. Then, the appropriate slice level Xi at each location is obtained by the same method (step d).

Then the light source A only illuminated (step e), to measure the slice level _{Z A i (i = 0,1,2 ...} ) for each window P ₀ to P ₄ at that time (step f), the following from them The coefficient C _A i (i = 0,1,2,3 ...) Representing the contribution of the light source A is calculated based on the equation (3) of (step g).

Here, the reference point P ₀ is set at the center of the display, but it is not necessarily the center because it is desirable.

Next, only the light source B is turned on (step h), and the coefficient C _B i (i = 0, 1, 2 ...) Is calculated in the same step (i, j).

In this way, the coefficients C _A i and C _B i in each window are determined as shown in FIG. 5, for example. As shown in the figure, the coefficient C _A i, C
_It can be seen that the value of _B i increases as it approaches the corresponding light source.

FIG. 4 is a flowchart of the process for the workpiece to be inspected, which is performed subsequent to the above-described pre-processing in the binarization processing method according to the present invention.

After the workpiece to be inspected is installed at a predetermined position (step a), the light sources A and B are lighted in the same manner as described above with reference to step c in FIG.
Slice level Y _'A, Y' representing the brightness of _B measured in (step b), a calculated binarization level correction value △ Xi in each window based on the following equation (step c).

_{△ Xi = (Y A -Y '} A) × C A i + (Y B -Y' B) × C B i
(4) Then, the proper slice level Xi obtained in the preprocessing is corrected by the calculated correction value ΔXi (step d).

〔The invention's effect〕

As described above, according to the present invention, it is possible to always set an appropriate binarization level even when illumination is performed by two or more light sources and the light sources have variations in brightness. Become.

[Brief description of drawings]

FIG. 1 is a block diagram of a system in which the method according to the present invention is implemented, FIG. 2 is a diagram for explaining a binarization process, and FIG. 3 is a flow chart of preprocessing in one embodiment of the present invention. FIG. 4 is a flowchart of a process following the preprocess in one embodiment of the present invention, and FIG. 5 is a diagram for explaining a method according to the present invention. In the figure, 10 ... Work, 12 ... Liquid crystal display, 14 ... Stopper, 16 ... CCD camera, 22 ... Personal computer, 100 ... Display segment, 102 ... Window.

Claims (1)

[Claims] 1. A plurality of measurement regions are defined on an image obtained by photographing a work illuminated by a plurality of light sources, and the brightness of a pixel in each measurement region is compared with a slice level set for each measurement region. In the binarization processing method of the image signal for binarizing the brightness of the pixel, (i) the brightness of each of the plurality of light sources from the image signal obtained by photographing the standard work illuminated by the light source of the standard brightness. To obtain a brightness measurement value for each standard light source, and (ii) a standard workpiece illuminated by only one light source of standard brightness The contribution of the brightness from each light source in each of the plurality of measurement areas is measured from the brightness in each of the plurality of measurement areas, and (iii) the same as the standard measurement of the brightness of each light source for the workpiece. Of each light source by (Iv) the difference between the standard brightness measurement value for each light source and the standard brightness measurement value for each light source, and the contribution of the brightness of each light source in each of the measurement areas A correction value for correcting the fluctuation in brightness is calculated for each measurement region, (v) the slice level is corrected by the correction value, and (vi) the brightness of the pixel in each measurement region is corrected by the corrected slice level. 2 of image signals characterized by binarization
Quantization processing method.