JPS63231580A - Data normalizing circuit - Google Patents

Abstract

PURPOSE:To obtain a data normalizing circuit normalizing a spatial filtering result by providing the data normalizing circuit with a positive data forming circuit finding out the absolute value of compression data outputted from a data compressing circuit. CONSTITUTION:The data compressing circuit forms compression data obtained by continuously extracting a part of data consisting of bits other than a sign bit to be the most significant bit of data processed by spatial filtering processing and adding a sign bit to the extracted data so that the sign bit is the most significant bit and applies the formed compression data to the positive data forming circuit. The positive data forming circuit calculates the absolute value of the compression data and outputs always positive compression data to a monitor. Thus, the spatial filtering result can be normalized without adopting a specified display system so that the spatial filtering result can be displayed on an ordinary monitor.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to a data normalization circuit for displaying the results of spatial filtering processing on image data on a normal gradation monitor.

(Conventional theater techniques) One of the conventional image processing techniques is spatial filtering processing. In this method, filter coefficients for the image data of a small area of nxn are subjected to a sum-of-products operation, and processing such as differentiation is performed depending on how the coefficients are given. usually,
The result of this product-sum operation is expressed by a larger number of bits than the number of bits of the input image data. Therefore, a monitor with very high gradation is required to display the image data after the spatial filtering. Also, image data is usually expressed as a positive integer, but the filter coefficient is a negative number (two's complement).
In some cases, the result of the product-sum operation becomes signed data (which means that it also takes negative numbers). Therefore, in order to use the product-sum calculation results as they are as image data, a special display method is required in terms of hardware, which complicates the device. However, such processing is limited by hardware and has the disadvantage of being impractical.

(Problems to be Solved by the Invention) Conventionally, when image data is spatially filtered, the result is represented by a larger number of bits than the number of bits of input image data, and may also be a negative integer. Therefore, in order to use the filtering results as they are as image data, a monitor with high gradation is required, and a special display method must be used in terms of hardware, which limits hardware processing and is not practical. There is a drawback that
Ta.

SUMMARY OF THE INVENTION Therefore, the present invention aims to eliminate the above-mentioned drawbacks, and provides a data normalization circuit that normalizes spatial filtering results so that the spatial filtering results can be displayed on a normal monitor without using a special display method. It is an object.

[Configuration of the Invention (Means for Solving Problems)] The present invention provides a system for spatially filtering image data, normalizing the result, and displaying the result on a monitor screen. a data compression circuit that creates compressed data by continuously extracting a part of data consisting of bits other than the sign bit and adding the sign bit to the extracted data so that the sign bit becomes the most significant bit; and a positive data generation circuit that takes the absolute value of the compressed data from the data compression circuit.

(Operation) In the data normalization circuit of the present invention, the data compression circuit continuously extracts a part of the data consisting of bits other than the sign bit of the most significant bit of the spatially filtered data, and converts the extracted data into Compressed data is created by adding the sign bit as the most significant bit, and this is supplied to a positive data creation circuit. The positive data generation circuit takes the absolute value of the compressed data and always outputs positive compressed data to the monitor.

(Example) An example of the present invention will be described below with reference to the drawings.
The figure is a block diagram showing one embodiment of the data normalization circuit of the present invention. 1 is a product-sum calculation circuit that performs a logical sum operation between a small area of input image data 100 and a filter coefficient; 2 is a product-sum calculation circuit that inverts the output data (20 bits) of the product-sum calculation circuit 1 (20 bits) 200; MSB processing circuit,
3 is a data select circuit that selects consecutive predetermined bits of data at a specified position from the output data 200 of the product-sum calculation circuit 1; 4 is a data selection circuit when the format of the output data of the product-sum calculation circuit 1 is a positive integer; Selects the MSB of the data output from the select circuit 3, and if the format of the output data is signed data (negative data), selects the MSB output from the MSB processing circuit 2 and outputs it. It is a selector.

Next, the operation of this embodiment will be explained. Below, as a general example, the input image data length is 8 bits (the data type is a positive integer), the filter coefficient length is 8 bits (the data type is a positive integer or signed data), and the filtering size is 3x3. Let me explain the case. In this case, the output data 200 after performing the product-sum calculation in the product-sum calculation circuit 1
has a maximum of 20 bits. If the filter coefficient is a positive integer, all 20 bits of the output data 200 will be data, but if the filter coefficient can be a negative number using two's complement, then as shown in FIG. As shown in A), among the 20 data pits, the MSB is the sign bit S. First, input image data 100 is subjected to a product-sum calculation with a filter coefficient in a product-sum calculation circuit 1, and output data 200, which is the result of the calculation, is output to a data selection circuit 3. At the same time, the MSB of the output data 200 is
Output circuit 2 is output to logic circuit 2.

The data select circuit 3 uses the sign bit S, which is the MSB of the input 20-bit data as shown in FIG. An arbitrary continuous bit string (the shaded part in FIG. 2(A)) is selected from among the 19-bit data, and a total of 8-bit data 300 is output as shown in FIG. 2(B). As a result, the output data 200 of the product-sum calculation circuit 1 becomes 8-bit data 300 in the data selection circuit 3, and its value is 128.
~127.

On the other hand, the MSI logic circuit 2 inverts the MSB of the input output data 200 and outputs the result to one input of the selector 4. Output data 30 is input to the other input of selector 4.
The MSB of 0 is input. Therefore, when the data format of the output data 200 of the product-sum operation circuit 1 is a positive integer, the selector 4 selects the MSB of the output data 300 as the output data 300.
If the data format of the output data 200 is signed data (negative integer), the inverted MSB of the MSI logic circuit 2 is output together with the remaining 7 bits of data.
It is output together with the remaining 7 bits of data of 00.

As a result, the spatial filtering results are normalized and unified into 8-bit positive integers as shown in FIG. 2(C), which is a normal image data format.

According to this embodiment, the output data (spatial filtering results) of the product-sum calculation circuit 1 can be normalized to an 8-bit positive integer, so the spatial filtering results can be displayed without any special display method. It can be displayed on a normal gradation monitor.

Furthermore, when the above normalized data is displayed on a monitor,
Although the absolute value of the spatial filtering result cannot be maintained, the relative value of each pixel in the spatial filtering result has meaning, and this relative value is maintained even when the above normalized data is displayed. It is possible to do this without compromising practicality.

[Effects of the Invention] As described above, according to the data normalization circuit of the present invention,
There is an advantage that the spatial filtering results can be normalized so that the spatial filtering results can be displayed on a normal monitor without using a special display method.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of a data normalization circuit of the present invention, and FIG. 2 is a diagram showing an example of data in each part of the circuit of FIG. 1... Product-sum calculation circuit 2... MSB processing circuit 3...
・Data selection circuit 4...Selector agent Patent attorney Noriyuki Chika Ken Yudo Sanno - Figure 1 (A) (B) (C) jI2 diagram

Claims (1)

[Claims] In a system that spatially filters image data, normalizes the results, and displays the result on a monitor screen, a part of the data consisting of bits other than the most significant sign bit of the spatially filtered data is continuously extracted. , a data compression circuit that creates compressed data by adding the sign bit to the extracted data such that the sign bit becomes the most significant bit; and a positive data creation circuit that takes the absolute value of the compressed data output from this data compression circuit. A data normalization circuit comprising a circuit.