JPH04184576A - Picture processor - Google Patents

Abstract

PURPOSE:To quickly obtain the cumulative value and mean value of the density values of arbitrary scattered areas by storing identification numbers divided at every area and given to each picture element of each area in each storing area corresponding to each picture element as they are. CONSTITUTION:Inputted video signals to be inspected are temporarily stored in a frame memory section 16 and binarized by using a prescribed density value as a threshold by means of the density converter 21 of a sector density adding section 17. A labeling section 18 groups adjacent picture elements which are lower or higher than the threshold of the binarized picture elements into areas, gives identification numbers divided at every area to each picture element in area, and stores each identification number in the look-up table 22 of the section 17 as it is through a looped data bus. Then the density converter 21 converts each picture element in each area to which the identification number is given into a multilevel density value and an adder 23 finds density sums at every area and calculates the mean value of the density values of each area by using the area value of each area to which the identification number is given. Therefore, the whole processes can be improved in speed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Field of Industrial Application) The present invention relates to an image processing device used for product inspection and the like.

(Prior Art) In recent years, with the advancement of image processing technology, inspection apparatuses have been developed that automatically perform product inspection, etc. without manual intervention.

Generally, in an image processing device used in such an inspection device, a video signal obtained by photographing a product using, for example, an ITV camera is processed into a plurality of sectors la, . . . , as shown in FIG.
In, each sector la.

..., In has the identification numbers "1" and "2°..." respectively.
and each identification number "1'" as shown in Figure 4.
, “2” . . . .

Prepare table 2 having 2o.

Then, each pixel in each sector la, . . . , In is converted into a multivalued density value, and the value is recorded in the memory area corresponding to the identification number.

As a result, the integrated value and average value of the density values of each sector can be determined and used for a predetermined inspection.

However, as shown in FIG. 5, when regions 3g, . . . , 3n for which the integrated value or average value of the density values is desired to be found are scattered arbitrarily, these cannot be determined by the above method.

For this reason, an image processing apparatus as shown in FIG. 6 has been proposed in the past.

The apparatus shown in the figure includes an image human power section 4. Frame memory section 5
．． Sector density adder 6. It is composed of a labeling section 7, a display section 8, and a CPU 9.

As shown in FIG. 7, the sector density adder 6 includes a density converter 10. Lookup table 11. Adder 12. It is composed of an address calculation circuit 13 and a result memory 14.

Then, the video signal of the object to be inspected (FIG. 8(a)) inputted from the image input M4 is temporarily stored in the frame memory section 5, and a predetermined density value is set for each pixel using the density converter 10. It is binarized as a value (FIG. 8(b)).

Next, the labeling unit 7 divides adjacent pixels of each binarized pixel below or above the threshold into regions, and each pixel in each region is given an identification number (label) "1" that is distinguished for each region.
”, “2”, etc. (FIG. 8(C)) are stored in the memory of the display unit 8.

Thereafter, the content (label image) of the memory of the CPU 9 or the display section 8 is read out for each address and stored in the look-up table 11 of the sector density addition section 6 (FIG. 8(d)).

Of the input image stored in the frame memory unit 5, each pixel in the area of the identification number stored in the lookup table 11 is converted into a multivalued density value by the density converter lO, and each pixel is converted into a multivalued density value for each area. Then, the adder 12 calculates the density addition value (FIG. 8(e)) and stores it in the result memory 14.

Furthermore, using the area value for each area of the identification number stored in the lookup table 11, the average value of the density value of each area is calculated.

(Problem to be Solved by the Invention) However, in such an image processing device, among the above-mentioned operations, the CPU 9 reads out the contents (label image) of the memory of the display unit 8 for each address and performs a look at the sector density addition unit 6. Since the operation of storing data in the up-table 11 takes a longer time than other processing, there is a problem in that the overall processing is extremely slow.

The present invention has been made based on the above-mentioned circumstances, and an object of the present invention is to provide an image processing apparatus that can quickly calculate the integrated value and average value of the density values of arbitrarily scattered areas.

[Structure of the Invention] (Means for Solving the Problems) The present invention provides a binarization means for binarizing a video signal of an object to be inspected using a predetermined density value as a threshold value for each pixel, and
Area identification means for dividing adjacent pixels below or above the threshold out of each pixel binarized by the digitization means into regions, and giving each pixel in each region an identification number that is distinguished for each region; A memory having a storage area corresponding to each pixel constituting the video signal of the object to be inspected, and storing each identification number of each pixel assigned by the area identification means as is in each storage area corresponding to each pixel. and density value calculation means for converting each pixel in the area of the identification number stored by the storage means into a multivalued density value and calculating the average value of the density values for each area. It is something.

(Function) In the present invention, since the identification number assigned to each pixel in each area, which is differentiated for each area, is stored as is in each storage area corresponding to each pixel, it is possible to store areas that are arbitrarily scattered. The integrated value and average value of concentration values can be quickly determined.

(Example) Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing the configuration of an image processing apparatus according to an embodiment of the present invention.

The device shown in the figure includes an image input section 15. Frame memory section 16. Sector density adder 17. The labeling section 182 and the display section 19 are connected in sequence, and each of these sections is controlled by the CPU 20.

Further, the output of the labeling section 18 is
The data bus is looped so that it is input to 7.

As shown in FIG. 2, the sector density adder 17 includes a density converter 21. Lookup table 22. Adder 23. It is composed of an address calculation circuit 24 and a result memory 25. Furthermore, the input to the sector density adder 17 is input to a density converter 21 and a lookup table 22.

Next, the operation of the image processing apparatus configured as described above will be explained.

First, a video signal of the object to be inspected inputted from the image input section 15 is stored in the frame memory section 16.

Next, each pixel is binarized using a density converter 21 using a predetermined density value as a threshold value.

After this, the labeling unit 18 divides adjacent pixels below or above a threshold value among each binarized pixel into regions, and assigns a unique number (label) to each pixel in each region, which is distinguished for each region.
"l", "2°, etc. are assigned to each pixel, and the individual numbers for each pixel are stored as they are in the look-up table 22 of the sector density adder 17 via a looped data bus.

Of the input image stored in the frame memory unit 16, each pixel in the area of the identification number stored in the lookup table 22 is converted into a multivalued density value by the density converter 21, and each pixel is converted into a multivalued density value for each area. Then, the adder 23 calculates the density addition value and stores it in the result memory 25. Furthermore, using the area value for each area of the identification number stored in the lookup table 22, the average value of the density value of each area is calculated.

In this way, in the image processing apparatus of this embodiment, the data bus is looped so that the output of the labeling section 18 is input to the sector density addition section 17, and each identification number of each pixel is connected to the looped data bus. Since the data is directly stored in the lookup table 22 of the sector density adder 17, this operation can be made faster like other processes, and as a result, the overall processing can be made faster.

[Effects of the Invention] As explained above, according to the present invention, the identification number assigned to each pixel in each area, which is distinguished for each area, is stored as is in each storage area corresponding to each pixel. So,
The integrated value and average value of the density values of arbitrarily scattered areas can be quickly determined.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of an image processing apparatus according to an embodiment of the present invention, FIG. 2 is a diagram showing the configuration of the sector density adding section shown in FIG. 1, and FIGS. 3 to 8 are FIG. 2 is a diagram for explaining a conventional example. 15... Image input section, 16... Frame memory section. 17... Sector density addition section, 18... Labeling section. 19... Display unit, 20... CPU, 21... Concentration converter. 22... Lookup table, 23... Adder,
24...Address calculation circuit, 25...Result memory. Applicant Toshiba Engineering Co., Ltd. Agent Patent Attorney Satoshi Suyama - (1 idiot) Figure 2 Figure 3 Figure 4 Figure 5 Figure 7 Figure 8

Claims (1)

[Claims] (1) Binarization means for binarizing the video signal of the object to be inspected using a predetermined density value as a threshold for each pixel; area identification means for dividing adjacent pixels below or above a threshold into areas and assigning each pixel in each area an identification number that is distinguished for each area; storage means having a storage area for storing each identification number of each pixel assigned by the area identification means as it is in each storage area corresponding to each pixel; An image processing apparatus characterized by comprising a density value calculation means for converting each pixel into a multivalued density value and calculating an average value of the density values for each region.