JPH03242772A - Digital image processor - Google Patents

Abstract

PURPOSE:To extract a feature value by a simple constitution by comparing an objective picture element(PE) and its near-by PEs with previously determined PEs, outputting data corresponding to the compared result as a bit string and using the data as an address signal to read out previously stored data indicating the feature value. CONSTITUTION:The digital image processor is provided with a comparing means 6 for comparing an objective PE and its near-by PEs with the previously determined PEs and outputting data corresponding to the compared result as a bit string and storage means 8, 9 for outputting data expressing features corresponding to the data in response to the reception of the bit string data as an address signal from the means 6. When the objective PE and its near-by PEs are compared with the previously determined PEs and the bit string data outputted correspondingly to the compared result are applied to the storage means 8, 9 as an address data, data expressing the feature of the objective PE are outputted in accordance with the states of the near-by PEs. Thus, the feature value can be extracted by the simple constitution.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a digital image processing device, and in particular, to digital image processing that extracts the number of boundary pixels, the number of connections, and the Euler number, which are local feature quantities of an image. PJ! Regarding equipment.

[Prior Art] Fig. 4 is a schematic block diagram showing a conventional digital image processing device, Fig. 5 is a diagram for explaining the feature amount of 3 x 3 pixels, and Fig. 6 is a diagram showing the Euler number. FIG. 3 is a diagram for explaining elements.

First, a conventional digital image processing apparatus will be described with reference to FIGS. 4 to 6. In image memory 1,
For example, 3x3 pixel X. -X8 is stored, and the feature amount is detected from the neighboring pattern of this image. This feature quantity is expressed by the number of connections, the number of boundary pixels, and the Euler number.

The number of connections is detected by the number of connections detection logic circuit 2,
The number of boundary pixels is detected by the boundary pixel number detection logic circuit 3, and the Euler number is detected by the Euler number detection logic circuit 4.

Here, the number of connections is the central pixel X. other pixel x1
It shows where x8 is located, and the number of boundary pixels is the center pixel X. Euler's number is the number of pixels adjacent to each other, and Euler's number is a powerful index that indicates the global characteristics of the shape of a figure, and can be used to find holes in a figure.

The Euler number is T shown in Figure 6(a) included in the figure,
It is obtained by summing the numbers of elements F shown in FIG. 6(b), E shown in FIG. 6(c), and E shown in FIG. 6(d). T
is an L-shaped three-pixel pattern and its axial symmetry as shown in FIG. 6(a).

It is a rotationally symmetrical pattern. F is a square four-pixel pattern as shown in Figure 6(b), and D is a pattern of four pixels as shown in Figure 6(b).
E is a two-pixel pattern arranged diagonally and its axially symmetrical pattern as shown in c), and E is a two-pixel vertically serial pattern or horizontally serial pattern as shown in FIG. 6(d). In the 3x3 pixel example shown in Figure 5, the number of boundary pixels is 4, the number of connections is 3, and the Euler number element T is 1.
, D is 3, E is 2, and F is 0.

[Problem to be Solved by the Invention 1] As mentioned above, the conventional digital image processing device has a logic circuit for detecting the number of connections, a logic circuit for detecting the number of boundary pixels, and a logic circuit for detecting the number of boundary pixels in order to separately detect the number of connections, the number of boundary pixels, and the Euler number. Since it is necessary to prepare dedicated hardware such as the logic circuit 3 and the Euler number detection logic circuit 4 and detect the respective feature amounts, the configuration has to become complicated.

Therefore, a main object of the present invention is to provide a digital image processing device that can output data representing the characteristics of a target pixel and its neighboring pixels with a relatively simple configuration.

[Means for Solving the Problems] The present invention is a digital image processing device that outputs data representing the characteristics of a target pixel and its neighboring pixels, and compares the target pixel and its neighboring pixels with predetermined pixels, Comparing means outputs data according to the comparison result as a bit string, and feature quantities that can be specified according to the states of neighboring pixels are stored in advance, and in response to the bit string data being given as an address signal from the comparing means. The apparatus is configured to include a storage means for outputting data representing characteristics corresponding to the data.

[Function] The digital image processing device according to the present invention compares a target pixel and its neighboring pixels with a predetermined pixel, and data of a bit string outputted according to the comparison result is given to the storage means as an address signal. , data representing characteristics is output according to the states of neighboring pixels.

[Embodiment of the invention] FIG. 1 is a schematic block diagram of an embodiment of the invention.
FIG. 2 is a concrete block diagram of the data conversion circuit shown in FIG. 1.

Next, with reference to FIGS. 1 and 2, the structure and operation of an embodiment of the present invention will be described. 3x3 pixel X. -x8 is stored as image data in the dual port memory 11 for one frame, for example, and is sequentially read out from the dual port memory 11 as one-dimensional image data. In order to extract neighboring feature quantities from this image data, it is necessary to collectively process 3×3 pixels. For this purpose, sequential image data is provided to the data conversion circuit 5.

As shown in FIG. 2, the data conversion circuit 5 includes D-type flip-flops 53 . 54 and 55.56
57, 58, and IH delay circuits 51 and 52 that delay one sequential line of data. The IH delay circuits 51 and 52 each delay sequentially input image data by one line, and the D type flip-flops 53 and 54 are connected to the IH delay circuits 51 and 52.
Stores image data delayed by one line each,
D-type flip-flops 55 and 56 are IH delay circuits 5
2, the image data delayed by one line is stored,
D-type flip-flops 57 and 58 latch directly input image data.

Therefore, among the sequentially inputted image data, pixels X4*X3+x2 are delayed by two lines by the IH delay circuits 51 and 52. Pixel X is output from the IH delay circuit 51;

is output from the D type flip-flop 53, and the pixel x
2 is output from the D type flip-flop 54. Also, among the image data, pixel Xs.

XQ and Xl are delayed by one line by the IH delay circuit 52, and the pixel X5 is output from the IH delay circuit 52 and becomes the pixel X. is output from the D-type flip-flop 55, and pixel X is output from the D-type flip-flop 56.

The image data converted into two-dimensional data by the data conversion circuit 5 is provided to a comparator 6 and compared with predetermined comparison data. The pixel value used as a reference for comparison is pixel X. pixel value or a specific pixel value consisting of . As a comparison output, the comparator 6 provides data 7 of a bit string in which po-po is arranged in the order of po, pp2, . . . po to the ROM 8.9 as an address signal.

In the ROM 8.9, a pattern in the vicinity of 3×3 pixels is used as an address, and feature quantities are stored in advance as data in accordance with the pattern in the vicinity.

That is, as shown in FIG. 3, for example, address r1
01011100J has the number of boundary pixels r0100J, the number of connections r011J rool indicating that Euler number element T is 1, rill indicating that J and D are 3, rlOJ indicating that E is 2, and F is O ``0'' indicating that ``0'' is stored. Therefore, comparator 6
Output 1) When the ROM 8.9 is accessed using o"po as the address, the feature amount at that time is output as data. The data of the feature amount output from the ROM 8.9 is selected by the multiplexer 10, and the necessary Note that data may be output for all pixels, not just a specific pixel, as a feature quantity.

C Effects of the Invention] As described above, according to the present invention, feature amounts that can be specified according to the states of neighboring pixels are stored in advance, and a target pixel and its neighboring pixels are compared with a predetermined pixel, Data corresponding to the comparison result is output as a bit string, and the data is used as an address signal to read data indicating the feature amount stored in advance, so the feature amount can be extracted with a simple configuration.

[Brief explanation of drawings]

FIG. 1 is a schematic block diagram of an embodiment of the present invention. FIG. 2 is a concrete block diagram of the data conversion circuit shown in FIG. 1. FIG. 3 is a diagram showing an example of data written to the ROM shown in FIG. 1. FIG. 4 is a schematic block diagram showing a conventional digital image processing device. Fifth
The figure is a diagram for explaining the feature amount of 3×3 pixels. FIG. 6 is a diagram for explaining Euler number elements. In the figure, 5 is a data conversion circuit, 6 is a comparator,
8.9 is ROM, 10 is multiplexer, 11 is dual port memory, 51.52 is IH delay circuit, 53-5
8 indicates a D type flip-flop.

Claims (1)

[Scope of Claims] A digital image processing device that outputs data representing the characteristics of a target pixel and its neighboring pixels, which compares the target pixel and its neighboring pixels with a predetermined pixel, and responds to the result of the comparison. a comparison means for outputting the data as a bit string, and a feature quantity that can be specified according to the state of the neighboring pixel, and in response to the bit string data being given as an address signal from the comparison means, A digital image processing device comprising storage means for outputting data representing corresponding features.