SU439000A1 - Device for reading images of objects - Google Patents

Description

I

The device is intended to read the image of a certain object from some given set of images for the purpose of further analysis and processing.

Device for reading images of objects are known, comprising a television sensor connected to a video signal extraction unit and a synchronization unit, sawtooth generators, counting pulse generators, counters, coincidence circuits, triggers, and a recording unit.

The known devices do not provide for the selection of the image of objects at a given raster location. In addition, they select images of objects with large errors.

The purpose of the invention is to improve the accuracy of the selection of an image of an object - the fact that the proposed device contains blocks for fixing the linear dimensions of objects in frames and rows, each of which consists of two identical channels containing a serially connected key associated with a video discriminator and a counting pulse generator, a filter counter, a digital-to-analog converter, a comparison circuit, the second input of which is connected to a sawtooth voltage generator, a one-shot connected to a key and

an exit trigger, where the trigger of the output of the block for fixing the linear dimensions of the object along the frame is connected with the keys of the block for fixing the linear dimensions for the rows, and the trigger for the output of the block for fixing the linear dimensions for rows is with the registration block.

FIG. 1 is a block diagram of the device; in fig. 2 - functional block diagram of fixing the linear dimensions of the object in rows.

The output of the television sensor 1 is connected to the discriminator of the video signal 2, the input is with the synchronizer 3. The output of the recorder 2 is connected to the output of the discriminator 2 and the inputs of the block fixing the linear dimensions of the images of objects 5 and 6, respectively, by frames and rows. The other inputs of blocks 5 and 6 are connected to the generator of counting pulses 7 and the corresponding saw-tooth voltage generators in frames and lines 8 and 9, the inputs of which are connected to the synchronizer 3. Blocks 5 and 6, which are connected mutually through the inputs and outputs, thereby form

sequential closed circuit. In addition, one of their outputs is connected to one of the inputs of the same block, as a result, in each block of fixing the linear dimensions of the images of objects, a reverse circuit is created

connection.

One of the outputs of the fixation unit in rows 6 is connected to the second input of the registration unit 4.

The device works as follows.

The video signal from the television sensor 1 is fed to the video signal discriminator, where it is clipped at a certain level selected according to the video signal processing algorithm. From the discriminator output, the video pulses are fed to the recording unit 4 and to the inputs of the block fixing the linear dimensions of the images of the objects in lines 6 and frames 5, where the sighting signal is compared with the corresponding sawtooth voltage on the rows and frames produced in the saw voltage generators 8 and 9. The generator of counting pulses 7 controls the inputs of fixing blocks 5 and 6, and the synchronizer 3 serves for accurate operation of the device as a whole.

Due to this scheme, not all images of objects that simultaneously fall within the field of view of a television sensor are analyzed and recorded, but one is selected, the location of which is defined by the parameters of fixation blocks 5 and 6.

FIG. 2 shows a functional diagram of the block fixing the linear dimensions of the image of objects 6. The functional diagram of block 5 is similar. Each of these blocks consists of two channels, fixing the values of the coordinates, -X in rows, -and frames, i.e., selecting the object with a rectangular reticle.

Each channel is a serial closed circuit consisting of a key 10, a counter filter 11, a digital-to-analog converter 12, a comparison circuit 13, and a one-shot 14. A discriminated video signal and counting pulses are supplied to the inputs of the key 10. The output of the one-shot 14 is connected to the input of the key 10 and the trigger 15.

The channels work as follows.

At one of the inputs of the comparison circuit 13, a sawtooth voltage is supplied from the generator 9, synchronized by television clock pulses (lines or frames), to the other - a constant voltage from the output of the digital-analog converter 12. At the moment of equal voltage on both inputs of the comparison circuit 13 At its output, a pulse appears that triggers a one-shot 14, which in turn generates a pulse of a fixed duration.

Thus, depending on the voltage at the output of the digital-to-analog converter 12, the pulse of the one-shot 14 moves along the television line. This pulse arrives at one of the inputs of key 10. The other input of the key receives a discriminated video signal, and the third receives counting pulses from generator 7. If the television line at the location of the pulse of the one-shot 14 has an object image, from the discriminator of video signal 2 to the input of key 10 impulse arrives, opening the key. In this case, the counting pulses of the generator 7 through the public key arrive at the counter-filter And, filtering the noise. In the event of a filter counter overflow, an overflow pulse is sent to the input of the digital-analogue to two converter 12, and the converter moves to the next discrete position. The voltage applied to the input of the comparison circuit 13 varies

one step, and its comparison with a sawtooth voltage occurs at the next level. This in turn causes a pulse offset of the one-shot 14 on the TV line. The described comparison process

flows over one television frame. In the next frame, the process is repeated, and, if the image of the object still coincides with the one-shot pulse, the D / A converter again moves to the next discrete position. This continues until the one-shot pulse comes out of the object image. The key 10 at the input of the discriminated video signal is closed, and the digital-to-analog converter 12 stops. The impulse of the one-shot 14 takes a fixed position relative to the image of the object, corresponding to its edge. In this way, automatic

object selection when reading the image.

Subject invention

Image reader

of objects containing a television sensor connected to a video signal discriminator and a synchronizer, sawtooth voltage generators, a generator of counting pulses, a recording unit connected to a video signal discriminator, characterized in that, in order to improve the accuracy of image selection of an object, it contains blocks for fixing linear dimensions of objects frame by frame and lines, including two identical channels in the form of serially connected keys associated with a video signal discriminator and a generator of countable pulses lsov, filter counter, digital-to-analog converter, the comparison circuit, the second

the input of which is connected to a sawtooth voltage generator, a one-shot connected to a key and a trigger, the trigger of the block for fixing the linear dimensions of the object along the frame is connected to the keys of the block for fixing linear dimensions in rows, and the trigger of the latter for the register.