JPH03128592A - Color picture signal evaluating method - Google Patents

Abstract

PURPOSE:To detect color shading of partial coloring, a zonal stripe of zonal coloring, or the like on a picture by dividing the picture to plural areas each of which has a unit area and counting the frequency in occurrence of the vector indicating color to obtain a histogram with respect to each divided area and detecting whether each area is colored or not. CONSTITUTION:The I signal and the Q signal stored in picture memories 3A and 3B are read out of same address as picture signals of the same picture element with respect to each divided area, and the coordinate position of the vector indicating color is stored in a coordinate memory 5 by vector operation in a computing element 4. The coordinate memory 5 is provided with storage areas 5A, 5B...5N whose number corresponds to the number of areas provided in picture memories 3A and 3B, and the frequency in occurrence of color is stored in coordinates (address) indicated by the vector indicating the color calculated by the vector computing element 4 in areas 5A to 5N. Histograms are taken in two directions of the I axis and the Q axis with respect 10 each of areas 5A to 5N of the coordinate memory 5 to specify the color having a er frequency in occurrence in accordance with two histograms.

Description

DETAILED DESCRIPTION OF THE INVENTION "Field of Industrial Application" The present invention relates to a color image signal evaluation method for evaluating a color image signal obtained from a color image sensor made of, for example, a semiconductor.

``Prior art'' A mosaic color filter or a stripe color filter is attached to the front of an image sensor made of a semiconductor.
Various color image sensors capable of obtaining color image signals have been put into practical use.

Image sensors made of semiconductors are manufactured using integrated circuit technology, but various defects occur due to defects in the manufacturing process. Especially when assembled as a color image sensor,
Defects in the image sensor cause color unevenness such as shading shown in FIG. 6 and band-like stripes shown in FIG. 7.

In other words, the shading shown in FIG. 6 refers to the color unevenness that occurs when a wide area of the screen (the figure shows the top left and bottom right areas colored) is colored on a screen that is supposed to be white.

Further, the band-shaped stripes shown in FIG. 7 refer to a phenomenon in which colored stripes appear diagonally or vertically on a screen that is supposed to be white.

Conventionally, color imaging signals were projected onto a color cathode ray tube,
This screen is visually monitored and detected.

``Problems to be Solved by the Invention'' Conventionally, color unevenness occurring in color image sensors has been artificially determined, which is inefficient. Particularly in mass-production factories, it is necessary to deploy a large number of inspectors, which is inconvenient as it goes against efforts to save labor and prevents cost reductions.

Furthermore, among defects, when color unevenness clearly appears, it can be detected without individual differences, but when color unevenness appears only faintly, it may be overlooked due to individual differences among inspectors.

An object of the present invention is to propose a color image signal evaluation method that can automatically detect color unevenness occurring in a color image signal obtained from a color image sensor or the like.

"Means for Solving the Problem" In this invention, an image on which a color image signal to be evaluated is projected is subdivided into a plurality of areas each having a unit area, and is applied to each pixel within the plurality of subdivided areas. The frequency of occurrence of the vector representing the color of the color signal is stored in a coordinate memory, and a histogram of the frequency of occurrence of the vector stored in this coordinate memory is obtained for each region. Determine the presence or absence of color unevenness.

According to the color image signal evaluation method of the present invention, since the frequency of occurrence of a vector representing the color of the color signal given to each pixel is determined for each subdivided region, the frequency of occurrence of each color can be totaled.

Furthermore, since the frequency of occurrence of vectors is determined for each area, it is possible to know the correspondence between positions on the screen and colors.

As a result, it is possible to detect color irregularities that occur locally, such as striped lines.

"Example" An embodiment of the present invention will be explained using FIG.

In FIG. 1, 1 indicates a color image signal source. This color image signal source 1 can be, for example, a color imaging device using a solid-state imaging device.

In this example, it is assumed that monochromatic signals R, G, and B are output separately from the signal source I, and these monochromatic signals R1G and B are input to the converter 2 and converted into an l signal and a Q signal.

The l signal, the Q signal, and the luminance signal Y are defined as follows, as is well known.

Y=0.3OR+0.59G+0.11B1=0.6O
R-0,28G-0,32BQ=0.21 R-0,5
2G+0.31 BR, G, and B indicate monochromatic signals of red, green, and blue, respectively, and each monochromatic signal R, G, and B is R=1. G
= 1° When B = 1, the luminance signal Y becomes Y = 1, and l
The signal is I=0. The Q signal becomes Q=0.

In other words, when the monochromatic signals R, G, and B are all l, a white image is displayed. The l signal and the Q signal are represented by the orthogonal coordinates shown in FIG. 2, and the color and 'a'a of the color are specified at each coordinate position on the orthogonal coordinates. The origin S represents white; when moving in the positive direction on the Q axis, the color becomes purple; when moving in the negative direction, the color becomes yellow-green; when moving in the positive direction on the I-axis, the color becomes orange; and when moving in the negative direction, the color becomes cyan.

In this invention, the screen on which the color image signal to be evaluated is projected is subdivided into a plurality of regions each having a unit area, and a vector representing the color of the color signal given to each pixel in the plurality of subdivided regions is generated. Store the frequency in coordinate memory.

For this purpose, image memories 3A and 38 are provided, each having a storage capacity corresponding to the number of pixels defined by the resolution of the color image signal to be evaluated.
0 image memories 3A and 3B that store the l signal and Q signal in
is divided into address areas corresponding to unit areas on the image, and reading is performed for each divided area.

The l signal and Q signal stored in the image memories 3A and 3B are read out as image signals of the same pixel from the same address for each partitioned area, vector-calculated by the arithmetic unit 4, and stored in the coordinate memory 5 as a vector representing the color. memorize the coordinate position of

The coordinate memory 5 is provided with a number of storage areas 5A, 5B, . . . , 5N corresponding to the areas provided in the image memories 3A and 3B, and vector calculations are performed on these areas 5A to 5N! The frequency of occurrence of the color is stored in the coordinates (address) indicated by the vector representing the color calculated in i4.

In other words, each area 5A to 5N provided in the coordinate memory 5 has an address space corresponding to orthogonal coordinates having orthogonal I and Q axes as shown in Fig. Write the number of times representing the frequency of color occurrence at the coordinates indicated by .

Therefore, if there are a plurality of pixels emitting the same color in the same area, the frequency of occurrence of that color is stored at the coordinate position indicating the same color.

When performing evaluation, it is common to provide a color image signal that displays a white screen, so the frequency of occurrence is concentrated near the origin S on the orthogonal coordinates representing white. 3 to 5 show vector occurrence frequencies on the I-Q coordinate axis and their histograms. In this example, each area 5A of the coordinate memory 5
This shows the case where a histogram is taken in two directions of the i-axis and the Q-axis every 5N. By taking a histogram over two dimensions in this way, it is possible to identify colors that occur frequently from the two histograms.

In the example of Fig. 3, the concentration is concentrated at the origin S, which indicates a good product, and in the example of Fig. 4, the concentration is concentrated at the origin S on the Q axis.
(2) There is scattering on the axis, and there is a tendency for color development to be observed, and it is determined to be defective.

In FIG. 5, the frequency of occurrence of vectors is scattered in both the I-axis direction and the Q-axis direction, and it is determined to be defective.

"Effects of the Invention" As explained above, according to the present invention, the screen is subdivided into areas having a plurality of unit areas, the frequency of occurrence of vectors representing colors is counted for each subdivided area, and a histogram is generated. Since this is determined, the presence or absence of coloring can be detected for each area.

As a result, it is possible to detect color unevenness where the screen is partially colored, band-like lines where the band is colored, and the like.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of this invention, FIG. 2 is a graph for explaining the relationship between color signals and vectors used in the embodiment of this invention, and FIGS. 3 to 5 Figures 6 and 7 are diagrams for explaining the coordinate memory used in the embodiment of this invention, the frequency of occurrence of vectors representing colors imported into the coordinate memory, and their histograms, and Figures 6 and 7 are diagrams for explaining types of color unevenness. FIG. 1: Signal source, 2: Converter, 3A, 38: Image mesori, : Arithmetic unit, : Coordinate memory, 5A to 5 N: SI area.

Claims (1)

[Claims] (1) The image on which the color image signal to be evaluated is projected is subdivided into multiple areas each having a unit area, and a vector representing the color of the color signal given to each pixel in the multiple subdivided areas is generated. A color image in which the frequency is stored in a coordinate memory, a histogram of the frequency of occurrence of the vector stored in the coordinate memory is obtained for each of the above regions, and the presence or absence of color unevenness on the screen is determined according to the form of the histogram in each region. Signal evaluation method.