JPS62254276A - Histogram calculation system - Google Patents

Abstract

PURPOSE:To simplify the constitution and also to increase the processing speed with a histogram calculation system by calculating approximately the overall histogram by applying the processing to only half the entire data. CONSTITUTION:A flip-flop 1 divides a pixel clock into half and switches the clock divided into half by gates 2-15 and its inverted clock in both ODD and EVEN fields. In the ODD field the data on the 1st, 3rd, 5th, 7th... columns are latched by a register 6. While in an EVEN field the data on the 2nd, 4th, 6th, 8th... columns are latched. The data latched by the register 6 are supplied to the addresses of a memory 8 and these data corresponding to addresses are read out of the memory 8 and latched by a register 9. Then the data latched by the register 9 are latched by a register 11 through an adder 10 which increases data by a piece and then written again to the original addresses.

Description

[Detailed Description of the Invention] [Prior Art and its Problems] Conventionally, in an image processing device, in order to obtain a histogram of the entire screen, it is necessary to calculate the number of each gray value for the gray data of each pixel on the screen. It is calculated by adding up.

This histogram is used for autofocus, autothreshold, and binarization, and needs to be processed at high speed. Therefore, it is particularly required to calculate this histogram in real time.

However, in the conventional method, the pixel clock frequency is high (approximately IQMHz), and the grayscale data is also input at this pixel rate, so the real-time histogram calculation circuit requires a large number of circuit configurations, or a very high-speed memory or IC. It becomes necessary to use .

[Purpose of the invention]

The present invention has been made in view of the above circumstances, and its purpose is to provide a method for approximately calculating the entire histogram by processing half of all data.

[Summary of the invention]

To facilitate understanding, let us now consider calculating a histogram of gray scale data on a screen as shown in FIG. The screen is composed of 512 x 512 pixels vertically and horizontally, and in the conventional method, processing is performed on data of 262,144 pixels of the entire screen.

On the other hand, with this method, data from every other column is stored in the first column of the screen, and data from columns not taken in the first column is stored in the second column. The histogram is calculated by integrating the data.

This is the same even if it is applied to pixels other than the shaded areas. In other words, any pixel may be processed.

Although the method of the present invention handles only 1/2 of the total data, since the data samples are spread uniformly over the entire area without depending on the direction, the overall histogram can be almost accurately calculated. It has the advantage of being able to approximate prices.

[Embodiments of the invention]

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

The second example shows an example of the N method for calculating the histogram of the screen in real time using the method of the present invention, in which 1 is a flip-flop that divides the pixel clock into 1/2. Note that this circuit assumes that the screen is scanned in an interlaced manner.

2 to 5 are circuits for switching the clock whose frequency is divided by 172 by the gate and its inverted clock between the ODD field and the EVEN field.

At the output of this logic circuit, the grayscale data inputted at the pixel clock rate is latched into the register 6.

The screen has ODD field and EVEN field as shown in Figure 4.
Every other line in the field is scanned.

In FIG. 3, grayscale data 1 to 10 indicate data for one column from the 1st column to the 10th column.

At this time, in the ODD field, 1°3.
5.7.9th column data is latched.

And in the EVEN field, 2.4°6.8
．． The data in the 10th column will be latched.

The data latched in the register 6 is input to the address of the memory 8, and the data corresponding to the address is read out from the memory 8 and latched in the register 9.

The data latched in the register 9 passes through an adder IO that increments the data by one, is latched in the register 11, and is written to the original address again at the cycle of the clock.

7 is a control circuit that generates a control signal for this circuit, and in this circuit, the data processed by the ODD field and EVEN field is in odd columns.

Since the columns are an even number, the data in the shaded portion in FIG. 1 will be processed as a whole.

〔Effect of the invention〕

As explained above, if the circuit is configured according to the method of the present invention,
Even if the pixel clock is high-speed, processing is performed using a clock with half the frequency of the pixel clock, which requires extremely high-speed memory and I/O.
Does not require C. Also, the circuit configuration can be made the simplest. Even if there is no need to calculate in real time, the data to be processed can be halved, so the processing time can naturally be cut in half compared to conventional methods.

[Brief explanation of drawings]

Fig. 1 is an explanatory diagram showing the structure of screen gradation data, Fig. 2 is a block diagram showing an example of a histogram calculation circuit according to the method of the present invention, Fig. 3 is a time chart of each signal, and Fig. 4 is a scan of the screen. It is an explanatory diagram showing a state. 1...Flip-flop 2...NOT gate 3.4...AND gate 5...OR gate 6...Register 7... ... Control circuit 8 ... Memory 9.11 ... Register 10 ... Addition circuit 12.13 ... Buffer No. 3'' Fig. 4

Claims (2)

[Claims] (1) In an image processing device that processes an image by A/D converting an image signal obtained from a camera, etc., in odd-numbered columns of the screen, the gradation data of the odd-numbered column is processed, and in the even-numbered column, the gradation data of the even-numbered column is processed. A histogram calculation method characterized by approximating the histogram of the entire screen by adding up the numbers for each density data. (2) In an image processing device that processes an image by A/D converting an image signal obtained from a camera, etc., in odd-numbered columns of the screen, the gray scale data of the even-numbered columns are processed; A histogram calculation method characterized in that the histogram of the entire screen is approximated by taking grayscale data and integrating the numbers for each grayscale data.