SU1462360A1 - Device for processing image signals - Google Patents

Abstract

The invention relates to information processing systems, in particular, to devices for processing and visualizing images. The aim of the invention is to improve the accuracy of the analysis of the object in its image by improving the conditions for the selection of low contrast details. The device contains a synchronous generator 1, an image signal sensor 2, a video monitoring unit 3, a digital-to-analog converter, an addresser 4, an image signal normalization unit 5, an analog-to-digital converter 6, interface 7 and 9 blocks, a control and calculation unit 8. automatic normalization of the extremes of the image signal, and, if necessary, its limitation with scaling, including adaptive, to improve the conditions of visual perception of the image of the observed object. 2 Il. with (L

Description

The invention relates to information processing systems, in particular to image processing and visualization devices.

The purpose of the invention is to increase the accuracy of the analysis of the object in its image by improving the conditions for the selection of low-contrast details.

Figure 1 presents a functional diagram explaining the operation of the device; figure 2 - plot characteristic signals and their histograms for various types of processing.

The device comprises a clock generator 1, a sensor 2 of the image signal, a video control unit 3, a digital-to-analog converter (DAC) 4, a block 5 for normalizing the image signal, an analog-to-digital converter (ADC) 6, a first interface unit 7, a control and calculation unit 8, and a second unit 9 pairing.

Figure 2 shows the input image signal 10 and its histogram 11, the image signal 12 and its histogram 13 after normalization, the output signal 14 of the image and its histogram 15 with restriction that is adaptive to the distribution of levels.

The device operates as follows.

Under the control of the synchro-generator | 1, the sensor 2 of the image signals and their receiver, for example, the video control unit 3, are synchronized.

In automatic normalization mode, the image signal 10 from the output of the sensor 2 is fed to the information input of the image signal normalization unit 5. In the time interval of the frame, the detectors of extrema (maximum and minimum), which are part of block 5, measure with digitalization of the maximum and minimum values of the image signal. During the frame backtrack, the measured values through the switches included in block 5 are overwritten in the corresponding extremum registers (maximum and minimum) also included in block 5. In the next frame, according to the analysis results under the influence of extremum codes .. recorded in the registers, the minimum value of the image signal to the black level (! '<,) ”and the maximum value to the white level (II g). The normalized image signal 12 is fed to the ADC input 6. The image signal is converted to a digital code synchronously with the clock frequency of the image elements, the pulses of which are fed from the clock 1 to the ADC clock input 6. The input dynamic range of the ADC 6 is matched to the dynamic range of the image signal from the output unit 5. The white level in the image signal corresponds to one, in all bits of the output of the ADC 6, and the black level - zeros. Such coordination allows reducing the requirement for ADC capacity to technically simple ones. From the output of the ADC 6, the signal can be input into a computer (not shown) for subsequent processing, or through the DAC 4 to the input of block 3 for image formation. Due to normalization, image reproduction is ensured over the entire range of possible brightness levels of the reproducing cathode ray tube of block 3.

If the image signal has rare amplitudes that are significant in amplitude (positive and / or negative), then normalization carried out on them does not lead to a noticeable improvement in the image, since the bulk of the details may appear in a narrow dynamic range of brightness, insufficient for reliable visual distinction. . In this case, improving the visual perception of the image can be achieved by limiting the extreme values of the image signal to the maximum and minimum threshold levels with the expansion of the contrast of the limited area to the nominal.

In digital image processing on a computer, it is necessary to take into account the possibility of increasing quantization errors in the processed signal, which are absent during processing in an analog device. The transfer of computers only to the function of controlling the levels of restriction “increases the accuracy of processing and unloads the computer for other tasks.

Let us consider the operation of the device in the controlled limitation mode for the case when the limitation levels are established based on the analysis of the statistical properties of the image signal. The normalized image signal is input to block 8 through the per- | vy block 7 pairing. The first pairing unit $ 7 can be made in the form of a controller for direct access to the operative memory of block 8 or include a buffer storage device for matching the information flow rate in the video path with the speed of the receiver (block 5). The operation of the first conjugation unit 7 is synchronized with a repetition rate of 15 image elements, and frame synchronization imgs are sent to it to identify the address of an element within the image field; pulses from the synchro generator 1. In order to ensure the possibility of connecting to block 8, in addition to the first pair of block 7 and other sources of information, the first block of block 7 should provide for a possible level of restriction to block 8, for example, in K% of the most common levels adopted in figure 2, position 13 for 100%. In the block, the values of N, and C ^ corresponding to the selected level of restriction are calculated. From the condition of geometric similarity of the histograms 11 and 13 of the abnormal and normalized signals 10 and 12, we find that “Tsmans - Umuw ^N 1 ----“ ------ And max

Tsmax - U mmn,,

^1, ma "s when using

U _O rp g ⁼ Umuh ⁺

255

8-bit ADC 6

The values of U _MWH and U _MaKC required for calculation are entered into block 8 from the corresponding detectors of block 5 through the second interface block 9.

The values of U _orp1 and U _orpi from block 8 through the second pair of block 9 and switches connected between the detectors and registers in block 5 neper disconnection of its outputs from the inputs of block 8, for example, by setting the outputs of the first block 7 of the pair in a high output resistance state . In block 8, the distribution of levels in the image signal is analyzed by calculating its histogram.

Figure 2 shows examples of the image signal 10 from the output of the sensor 2, where is the maximum possible value of this signal by its histogram .11, The abscissa axis shows the values of the input signal, and the ordinate axis shows the frequency of occurrence of these values in the signal during the frame time. Position 12 displays the same signal after normalization, and its histogram ^ calculated by the formula "> -,

N = 0 where N is the number of the image signal level (it is accepted that the signal is converted into an 8-bit code, i.e., N = 0 - 255);

η “the number of image elements in the frame corresponding to the image signal level Ν is indicated by 13.

To calculate the input signal restriction levels and _op1 and U _orp2, at the request of block 8, the operator enters into the register in the corresponding registers. To the levels of black and white at the output of block 5 (Fig. 2, position 14), not the minimum and maximum values of the input signal are now tied, but its values corresponding to both _ogre and U _Qrp2 . A histogram of such a signal is presented at position 15, Fig.2. This improves the visualization of the image sections corresponding to the average brightness levels. Signal values below the black level and above the white level are limited in the ADC 6, due to the limitation

4 ₀ maximum values in terms of white on the screen of unit 3 reduces the exposure from rare bright details, which reduce the overall contrast.

The amplitude characteristic of the discharge path takes the form I

U ₆ At U _8х (t) ⁷ / U _O rp2

A -U _BX (t) for U _orp . _{-7U & x} (t) ⁷ U _Or pf. U ₄ for U (t) SU _orpi and € _ _Ξ _υ + ___

Uorp-1 - Pogrt where A =

Thus, the second pairing unit 9 should provide four directions for transmitting information: two 55 from the data output of block 8 (to the inputs of the maximum and minimum registers of block 5) and two to the data input of block 8 (from the detectors of maximum and minimum of block 5). The transmission direction is set programmatically, from block 8 and is accompanied by the installation of the corresponding code on the address output of block 8 ‘. To this end, the second interface unit 9 must include switches and an address decoder. If the unit 8 does not access the second interface unit 9, its data outputs are transferred to the HIGH OUTPUT resistance 1Q state.

Thus, the described device can automatically normalize the extremes of the image signal, and if necessary .. 15th | restriction with scaling, including [adaptive, to improve the conditions for visual perception of the image of the observed object.

As a consequence of this, by setting different levels of restriction, it is possible to stretch individual sections of the signal amplitudes to the entire dynamic range of the reproducing device (method of amplitude cuts).

The amplitude processing in the described device is superior in efficiency to that performed in a computer over a digital signal, since it does not increase the quantization noise in the processed signal, which also improves image analysis c; low contrast details. The processor of block 8 is used only to calculate and set the parameters of the restriction mode and free from the task of element-wise conversion of the image signal. At the same time, during processing, block 8 can perform other tasks related, for example, to monitoring system parameters.

In those cases when the essential information is contained not only in the spatial characteristics of the observed object, for example, when using the device in a television system with a scanning electron microscope _? possible indication or logging on the peripheral devices of block 8 of the absolute values of the extrema of the image signal, input from block 5. This eliminates the difficulties in comparing the properties of the observed objects with different reflection characteristics of the radiation incident on them, the images of which are displayed on the screen of block 3 due to processing in the same range of brightness.

Claims (1)

Formula of isobrtea A device for processing image signals containing a clock, a normalization block for the image signal, the synchronization input of which is connected to the output of the frame clock of the clock, the information input of the block for the normalization of the image signal is the information input of the device, characterized in that, in order to increase the accuracy of the analysis of the object by image by improving the conditions for the selection of low-contrast parts, the first and second interface units, an analog-to-digital converter are introduced into it 25 a control and calculation unit, wherein the output of the pulse pulses of the image elements of the clock generator is connected to the synchronization input of the analog-to-digital converter 30 and the input of the pulse pulses of the image elements of the first coupler, the input of the frame sync pulses of which is connected to the synchronization input of the normalization unit __ of the image signal, the first information output of which is connected to the information input of an analog-to-digital converter, the output of which is connected to the information input 4Q of the first unit Since the interface is the information output of the device, the information output and the synchronization output of the first interface unit are connected to the inputs of the same name 45 of the control and calculation unit, output. data, the address and synchronization of which are connected to the same inputs of the second interface unit, the first and second inputs of the extrema ₅ 0 of the second interface unit are connected to the same outputs of the normalization block of the image signal, the first and second inputs of the extrema and the mode input of which are connected to 55 of the same output ⁴ the second interface unit, the data output of which is connected to the information input of the control and calculation unit.