JPS60221877A - Arithmetic unit of figure - Google Patents

Abstract

PURPOSE:To accelerate substantially checking of overlapping of figures by holding coordinate position information of two and more rectangles and by performing simultaneously the figure arithmetic with respect to each figure data by means of a full function memory element. CONSTITUTION:Plural function memory elements 711-7nn are constituted of two pieces of one bit memory elements 16a and 16b and one bit arithmetic unit 17 as a main component. X and Y address control circuits 5 and 6 input addresses of two points on coordinates, output an address selection signal with respect to the whole area held by coordinates, while X and Y direction registers 8 and 9 hold an OR summed up by contents of the memory elements 711-7nn arrayed two-dimensionally in row and column directions. Coordinate data is written in the memory elements of the function memory elements selected by the circuits 5 and 6 out of the memory elements 711-7nn, and coordainte position information of two and more rectangles is held. The figure arithmetic is simultaneously performed with the aid of a full function memory element.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] 1. The present invention relates to a graphic calculation device that performs calculations on a graphic whose individual graphic elements are represented by rectangles.

[Prior art]

A conventional graphic arithmetic device of this type is shown in FIG. In the following, a graphic arithmetic device that determines whether or not a plurality of rectangles have mutually negative portions in a graphic arithmetic operation will be described as an example. In FIG. 1, 11 to 14 each represent a rectangle l.
21 to 24 are comparators that input two coordinate values and determine their magnitude; 31 to 34 are the comparators 21; -24 is a comparison output indicating the comparison result, and 4 is a multi-a flag obtained by taking the AND of all the comparison outputs 31-34.

Next, the operation shown in FIG. 1 will be explained below. Here, it is assumed that each solid element is a rectangle surrounded by CX-y) II marks at the start point and the end point, and the x and yMk marks at the end point are each larger than the x and y coordinates of the start point. When calculating whether there is an overlapping portion between two rectangles, rectangle 1
The (x-y) coordinates of the starting point and ending point of rectangle 2 are also stored in registers 11 and 12.

The (x-y) coordinates of the end point are stored in register 13.14. The conditions for rectangle l and rectangle 2 to have a quadrature square part are that the 1st and 7th coordinates of the end point of rectangle 2 are larger than the X and 7 coordinates of the start point of rectangle l, respectively, and the end point of rectangle l is The 1st and 7th coordinates are also larger than the x and y coordinates of the starting point of rectangle 2, respectively.

In order to confirm the position of each of the above X, 7 coordinates by calculation, 4
The starting point and ending point of each rectangle are input to the comparators 21 to 24! ,7
Input coordinates and perform comparison calculations.

As a result, if the above conditions are satisfied, each comparison output 31
34 respectively indicate a true state, and the logical product of the four comparison outputs 31 to 34 is set as the overlap flag 4, it is determined whether there is an overlap between rectangle 1 and rectangle 2 based on the truth or falsehood of this overlap flag. It will be possible.

As described above, in conventional graphic arithmetic devices, in order to determine whether two rectangles overlap, each starting point is used.

Input the (x-y) coordinates of the end point into the register, perform four comparison operations, and summarize the results to determine if the rectangles overlap. In order to determine whether or not there is, it is necessary to perform the above-mentioned determination process seven times, and when the value of the rectangular data is large, a huge amount of determination time is required.

[Summary of the invention]

This invention was made in order to eliminate the drawbacks of the conventional ones as described above, and has two two-dimensional array memories each having a storage capacity of one pit. y) By adding the function of writing to the memory part corresponding to the coordinates by giving the coordinates and the bit processing capacity for the corresponding l bits of two two-dimensional array arrays, it is possible to write between graphic elements. It is an object of the present invention to provide a graphic arithmetic device that checks 'xLa and performs various graphic arithmetic operations.

[Embodiments of the invention]

An embodiment of the present invention will be described below with reference to the drawings.

In FIG. 2, the same parts as in FIG.
A register that stores the X coordinate of the end point and one coordinate of the start point and the end point; 5 is an X address control circuit that inputs two X coordinates and outputs an X address selection signal for the area sandwiched between the two X coordinate values; 6 is a Y address control circuit that inputs two y coordinates and outputs a Y address selection signal for the area sandwiched between the two 1 coordinate values; 111 to 7nn;
Hiro functional memory element, 8-mine functional memory element 711~''7n
X to hold data propagated in the X direction from n
A direction register 9 is a Y direction register for holding data propagated in the Y direction from the functional elements 111 to 7nn, and 10 is a memory selection for selecting either of the two storage elements in the functional memory. Line, 1lla~11n&
111b is an X-person address selection line for determining the X-direction address of the A-side memory of the storage element in the functional memory;
~11mb is an XB address selection line for determining the X-direction address of the B-side memory of the storage element in the functional memory, and 121-12m is a Y address for determining the Y-direction address of the storage element in the functional memory. Selection line, 1301-1
3El11 is an X data propagation line for transmitting data from memory elements 711 to 7nn in the Y direction, 1410 to 1
wa■ is the direction from the functional memory elements 111 to 7nn in the X direction.
-Y data propagation line 15A is a function designation line for controlling the function memory 711 to 7nn (DhJJ production). , 11 and 16b are storage elements each having one pit, 171i is a bit arithmetic unit that accepts two 1-bit inputs, performs various operations and generates a 1-bit output, 1lia is an XA address selection line, and 111b is an XB Address selection green, 12jaY address selection line, 13(1-
1) J, 131j is the X data propagation line, 141 (J-1
), 141J is a Y data propagation line, and 151 to 154 are function designation lines, which are a reset gland two-person write line, an 8-convolution line, and an arithmetic function designation line, respectively.

Next, the operation of the graphic arithmetic device shown in the second embodiment and FIG. 3 will be explained. Here, each graphic element is represented by the (x-y) coordinates of the starting point and ending point of the rectangle, and its X.

It is assumed that the y value takes a value from 1 to nt corresponding to the number of functional memory collections included in the graphic processing device. Prior to the start of operation, two storage elements 16a, 16b among the full-function memory elements 711-7nn are activated by the reset line 151.
Reset everything. When calculating whether there is an overlapping shearing part between two rectangles, first store the (x-y) coordinates of the starting point and ending point for the first rectangle in register 15.
．． 16. These x, y coordinate values are transferred to the X address control circuit 5 and Y address control circuit 6, respectively, and the XA, XB, and Y address selection lines 111a corresponding to the areas sandwiched between the starting point and ending point of each x, y coordinate are transferred. ~1l
na t 111b ~ l1nb + 121 ~ 12n
takes a true value. However, at this time, the X address control circuit 5
Depending on the value of the memory selection line 1o transferred to the Since the functional memory elements 111 to 7nn are connected by the same signal line, all the functional memory elements 71J corresponding to the coordinates of the input rectangular start and end points are connected to the address selection k in the X direction. 111a to 11na or 111
b ~ 11nb and the Y address selection lines 121 ~ 12n are both true values, and as a result, the AND circuit for the address selection lines in the functional memory Tlj selects either one of the two 1-pit storage elements 16a, 15b. is set. For example, the coordinates of the starting point and ending point of a rectangle are (3°s) +
When the memory selection line 10 specifies the A side in (617), 12 sets of functional memory elements 735, 736° 737',
745,746,747,755,756°757.7
65,766.767 are the above XA and XB.

The memory element 1e of Altll among the functional memory elements 135 to 767 is selected by each address selection line of Y.
a is set. Next, the coordinates of the start point and end point of the second rectangle are similarly input into the registers 15 and 16, and all functional memory elements 71J corresponding to these coordinates are selected. At this time, by setting a value different from the first value to the memory selection line 10, it is possible to set the unused memory element 16b among the functional memory elements 71j.

As a result of the above operations, the figure corresponding to the rectangle 1 is stored in the functional memory element 71j, the OA side storage element 16&, and for the other rectangle 2, ii, and the B side storage element 16b, in a bit-corresponding manner. Element 11 for all functions of the device
For 1 to 7nn, logical product is applied to the arithmetic function designation line 154,
Specify true for the Big-blind line and set the functional memory index '111
By operating the ~7nn pit calculator 11, A
The result of the logical product of the contents of the side storage element 16a and the contents of the B111l storage element 16b is set in the B side storage element 16b.

If the two rectangles overlap at a specific point due to this operation, it means that both memory elements 16&, 16b of the functional memory element 71j corresponding to that point are set to true. As a result of performing the above calculation, Bll
! The true value of 16 elements of the l storage element will be held.

Therefore, by checking the contents of the B side storage elements 16b of the full-function memory elements 111 to 7nn, it is possible to check the overlapping results of the rectangular diagrams. For this purpose, all functional memory elements 711 to 7 of this graphic arithmetic unit are used as the starting and ending points of the rectangle.
1 and n in register 15 to specify nn.

By setting 0 and 0 in the register 16 and instructing the memory selection line 10 to specify the B side, all the XB address selection lines 111b to 11nb become true.

At this time, if all the endmost X data propagation lines 1301 to 130n are set to a false state, each functional memory element 71
Inside j, the X data propagation line 13(1-1)j
Since the logical sum of the contents of the B-side storage element 16b and the contents of the B-side storage element 16b is taken and outputted as the X data propagation line 131J, the final X data propagation lines 1113nl to 13nn have full-function memory collections 111 to 111 for each X direction. The sum of B stu storage element b(2)Ltj of 7an is obtained, and this result is set in the X direction register 8. Therefore, by determining whether the contents of this X-direction register 8 are all false, it is possible to determine whether two rectangles have one or more copies. If there is even one item that is not false, the cows that have duplicated parts in the corresponding column are indicated, so in this case, as with the data propagation in the X direction, register 16 is By setting 0 to the register 15 and an even state to the end Y data propagation lines 1410 to 14nO, the contents of the B side storage element 16b of the full function memory elements 711 to 7nn are logically stored in each X direction. The sum is calculated and the result is set in the Y-direction register 8. Therefore, X
It turns out that there is an overlapping portion at the location specified by the direction register 8 and the Y direction register 9.

When there is no overlap between two rectangles and a third rectangle is checked for overlap between these two rectangles, in addition to the A-side storage element 16m of the functional memory element 71j that holds rectangle l, Rectangle 20 figure data is also this A II
Set in I memory element 16m. Therefore, A side storage element 1
6m holds the graphic data of both rectangle 1 and rectangle 2.

At this time, data of rectangle 3 is stored in B of functional memory element 71J.
By setting the data in the fIQ storage element 16b and performing the same calculation as described above, a rectangle l is created.

The rectangle 30 overlap test for 2 is performed in one operation. Thereafter, the TL double check with the next rectangle may be performed while the rectangular data determined to have no xi are sequentially added to the functional memory element 7ijOA side storage element.

In the above embodiment, the structure of the functional mesori element 71j is n.
Although it is arranged in rows and columns, a configuration of any size is possible, and even if this configuration does not directly correspond to the (x-y) coordinate values of each rectangular data, the x, y address control circuits 5 and 6 1 by adding, subtracting, multiplying or dividing by an appropriate constant.
You can also cupping. Furthermore, when storing each rectangular data in the memory elements 16m and 16b of the functional memory element T1j,
Expand, reduce, and shift this rectangle using an appropriate constant.

The address selection line may be specified based on the result of calculations for X-axis, Y-axis, etc. Further, in the above embodiment, only the logical product is shown as the bit operation device of the functional memory element 71j, but various operational functions such as logical sum, negation, and exclusive logical sum may be added, and the calculation result can be temporarily used. A register to hold the data may be added.

〔Effect of the invention〕

As described above, the present invention provides a graphic arithmetic device in which functional memories including two storage elements and a bit arithmetic unit are arranged two-dimensionally, and the (x-y) 411 was created by adding a circuit that selects all functional memories included in the range from the coordinate value, so the duplicate check for figures that conventionally required seven times for each figure data was reduced to k times. This has the effect of greatly speeding up the process, which is possible with return.

[Brief explanation of the drawing]

Figure ta1 is a configuration diagram showing a conventional solid-state arithmetic device, and FIG. 2 is an overall configuration diagram of a graphic arithmetic device according to an embodiment of the present invention.
FIG. 3 is a circuit diagram of a functional memory used in this embodiment. 5...X address control circuit, 6...Y address control circuit, 111-7nn...functional memory element, 8 ...
X direction register, S...Y direction register, 10...
Memory selection i, 16&, 16b...Storage element, 17.
...1 pit calculation unit. Patent applicant Mitsubishi Electric Corporation

Claims (1)

[Scope of Claims] A plurality of functional memory elements configured with two 1-bit storage elements and a 1-bit arithmetic unit as main elements. an X-address control circuit and a Y-address control circuit that input address values of two points on coordinates and output address selection signals for the entire area spread between the coordinate values; and the plurality of points arranged two-dimensionally. The two-dimensionally arranged function has an X-direction register and a Y-direction register that hold the result of ORing the contents of the memory element in the row direction and the column direction, and the function is arranged two-dimensionally. Among the memory elements, coordinate position information of two or more rectangles can be stored by convolving the coordinate data into the memory element of the functional memory element selected from both the X address control circuit and the Y address control circuit A. A graphic arithmetic device having a fist as a symbol, and capable of simultaneously performing graphic arithmetic operations on each one shape data using the full-function memory element.