CZ309660B6 - A method of measuring the distribution of dust particles in space and time and a device for this - Google Patents

Abstract

A method of measuring the distribution of dust particles in space and time, where the following steps are carried out sequentially:a) a gaseous medium is forced into the measured space,b) the space is illuminated by a laser beam,c) a digital image of laser-illuminated dust particles is created,d) the digital image is converted into matrix form,e) background correction for systematic errors is carried out,f) thresholding and binarization of the matrix form of the image is performed,g) the matrix form of the image is searched and the areas corresponding to the illuminated dust particles are searched,h) the size of the found areas is calculated ai) the calculated values are transferred to the database of dust particle sizes.A device for carrying this out which consists of a circulating fan (2), a laser (3) and a digital camera (4) with optics, an evaluation device (5) arranged one behind the other, the background correction algorithms are integrated in the evaluation device (5) for systematic errors, for thresholding and binarization of the image in matrix form, for searching the matrix form of the image and finding areas corresponding to illuminated dust particles and conversion to the database of dust particle sizes. The digital camera (4) is bi-directionally connected to the evaluation device (5).

Description

A method of measuring the distribution of dust particles in space and time and a device for carrying out this method

Field of technology

The invention falls into the field of digital image processing and the field of sensors.

Current state of the art

Equipment for digital image evaluation, or there are many types of devices that process digital images for their purpose. It is most often possible to meet them in an industrial environment, when such equipment evaluates, for example, the surface of products and looks for surface defects, checks the diameter of holes, the position of grooves, the depth of cuts, the resulting color of the products, etc. The goal is usually to replace the human factor during the repetitive evaluation of one of the indicators product quality. Machine evaluation is generally more objective, independent of the time of day and the number of hours worked. Unfortunately, the machine evaluation is objective only and only if it works in constant working conditions, which are as close as possible to the conditions during machine calibration. In the event of a change, for example, in the intensity of the ambient light, the brightness threshold set internally by the algorithm may shift and the device will start, for example, to ignore defects or generate false detections, etc. Such conditions must then be validated manually, and the human factor is again involved in the quality assessment process. These conditions are of course unwanted and make production more expensive and slower. In the case of using devices for quality control using digital image evaluation, the goal is always to achieve the greatest possible robustness not only of the mechanical elements, but also of the control and evaluation algorithm. The mentioned method of image processing is described, for example, in patent CZ 308988.

A special application in the mentioned area is the detection of dust particles. Nowadays, dust particles are in the area of interest both from the point of view of the environment and from the point of view of the health impact on the population. Dust particles (also known as PM, from English particulate matter) are categorized according to their size, because it is their size that has a direct impact on health.

The main components of PM are sulfates, nitrates, ammonia, sodium chloride, black carbon, mineral dust and water. Particles with a diameter of 10 pm or smaller are the most harmful from the point of view of the effect on human health, because such small particles can penetrate through the lungs into the bloodstream. The World Health Organization (WHO) has designated airborne PM as a group 1 carcinogen. Studies to date have shown that there is no safe level of exposure to PM, and with any increase in the concentration of PM in the air, the incidence of cancer increases proportionally.

It is usually possible to meet dust particle sensors detecting particles in the sizes PM1.0, PM2.5, PM4, PM10, etc. This corresponds to the absolute particle sizes of 0.3 to 1.0 pm, 0.3 to 2.5 pm, 0 .3 to 4.0 pm, .3 to 10 pm, etc.

Individual sensors detect PM in a given size interval, or they detect the number of particles of a given size in a certain time. If it is necessary to detect the sizes of each individual particle, it is necessary to group a large number of different sensors and compare their outputs. The inventions described, for example, in patent applications JPH 09243547 and JP 2000214071 are based on this principle, where JPH 09243547 describes a method of detection by illuminating a liquid/gas with a laser beam to create diffraction scattering on dust particles. Scattered light is focused using a front and rear lens system into an optical preamplifier, amplifier and detector. The analog detector is digitized for numerical data processing. JP 2000214071 is also based on light scattering of a laser beam on a dust particle, but uses slit lens arrays and a single photodetector in the form of a photodiode with optical filters. The scattered light generates a current in the photodiode. The number of current pulses corresponds to the number of dust particles passing through the tunnel/sensor. Another solution is described in the application

- 1 CZ 309660 B6

JPH 09101252, where the principle of diffraction of light on dust particles is used. The laser beam is concentrated using a lens into a coherent beam. Upon impact on a dust particle, the light will be diffracted, and the lens behind the area of the dust particles will focus the diffraction on the detector. It consists of circular optical detectors that generate information about a dust particle.

The mentioned methods are not very efficient and introduce errors into the measurement chain. On PMs, if they are illuminated, the light falling on them is scattered. The properties of light scattered on particles are used by sensors to measure their number, size and concentration. The basic components are a light source directed to the detection area in the detection chamber, a detector measuring the light scattered by the particle, and electronic circuits that process and analyze the output from the detector.

The sensors include a circulation fan, a tunnel for conducting a gaseous medium containing dust particles, a detection chamber, an electronic evaluation device and peripherals for communication with the environment.

The circulating fan is an electromechanical component that ensures a constant (as much as possible) gas flow over time. The circulation fan determines how much media passes through the sensor.

The detection chamber is an area inside the sensor that is usually darkened and contains a light source, usually a laser beam, that illuminates the gaseous medium. Its purpose is to shield external influences and thus increase the distance between the signal and the useful signal.

The evaluation device is usually implemented with electronic circuits that measure the intensity and length of the light flash generated by a dust particle (PM), illuminated by an intense light source (LED, LASER). Its purpose is to evaluate and present measurement results to an external device.

The essence of the invention

The essence of the invention is the measurement of the distribution of dust particles in space and time and the device for performing this method. The device (Figure 1) contains an opening for the entry of a gaseous medium containing dust particles, a circulation fan, a line, or a tunnel for channeling the medium, a detection chamber with a laser beam, a digital camera with optics, an evaluation device based on a personal computer and a database of evaluated data, or frequency of occurrence of dust particles over time and their size. In this invention, the term "tunnel" refers to the hollow part of the device through which the gaseous medium passes from the entrance to the device, through the fan and the detection chamber to the exit.

The circulating fan ensures a constant flow of gas medium through the device. It is possible to control it with the aim of influencing the flow rate of the medium, or its flow volume. The nature of the medium flow through the tunnel is predominantly turbulent behind the circulation fan, gradually becoming predominantly laminar.

The essence of the method is to perform the following steps:

a) driving the gaseous medium into the measuring space,

b) illumination of the space with a laser beam,

c) creation of a digital image of laser-illuminated dust particles,

d) conversion of a digital image into matrix form,

e) performing background correction for systematic errors,

f) performing thresholding and binarization of the matrix form of the image, g searching the matrix form of the image and finding the area corresponding to the illuminated dust particles,

h) performing a calculation of the size of the found areas a

i) transfer of calculated values to the dust particle size database.

In a darkened detection chamber, the medium passes through a laser beam, which if in the medium

- 2 CZ 309660 B6 are not dust particles, the space in the detection chamber does not illuminate in any way, and is therefore not visible. If a dust particle is present in the medium, it is illuminated by the laser beam. After being illuminated, the particles reflect part of the incident light, resulting in an intense glow. This reflected light is captured by a digital camera that continuously scans the area in the detection chamber.

The digital image (Figure 2) captured by the camera is converted into a two-dimensional matrix, where each cell of the matrix corresponds to one pixel of image information. The content of the cell is a number in the range 0 to 255, which corresponds to the brightness of the given pixel. A pixel without illumination shows a brightness value of 0, a fully illuminated pixel shows a value of 255, so-called saturation. There are other digital image sensors that have a maximum pixel brightness value other than 255. However, this is of no practical importance for the purposes of this invention. In this context, it is therefore possible to confuse the term matrix cell and image pixel. A matrix cell physically represents a real image pixel of the sensor. The matrix mostly contains only dark pixels, only in places where a dust particle is detected, the pixels have a higher brightness value.

The digital image is binarized (Figure 3), or the values of the cells of the matrix are converted to the values 0 or 1, respectively. the possible range of values in a cell is limited to only two values. Binarization is the process where each cell of a matrix is compared to a user-defined threshold value and assigned a new value. This converts an image in shades of gray (pixels take on values from 0 to 255) into a black and white image. This eliminates image digitization errors, various reflections caused by reflections in the detection chamber, etc. Dust particles captured by the camera stand out, forming clearly defined clusters of cells with values greater than the values of cells representing dark areas.

Furthermore, background correction for systematic errors is performed. It is possible that, for example, a digital camera contains defects that are manifested by the appearance of pixels with a static, unchanging brightness value (so-called defective pixels). These defective pixels are always in the same place and consistently give false information about the condition of the captured image. It may also happen that there is a dust particle in the scanned area that has stuck to the wall and reflects the light. Again, this is usually a fixed defect that does not correspond to the actual state of affairs in the detection chamber. These continuous defects need to be removed by comparing the current image captured by the digital camera with a reference image where the continuous defects are captured. The detection area is then limited to areas outside of the systematic error areas. In this way, most false detections are eliminated and the image corresponds to the actual state of the gas medium in the detection chamber.

Next, a dust particle in the image is found. This is to find the beginning (Figure 4) of a suspicious area that corresponds to a cluster of bright pixels on a dark background. The entire image is searched, or matrix, and when the first bright pixel is found, the search is stopped. A cluster size estimation algorithm is run, where the default detection coordinate corresponds to the coordinate of the first cluster pixel found in the matrix. The output of the algorithm is the size of the square area in which the found cluster of bright pixels is located.

In this area, the number of bright pixels corresponding to the size of the captured dust particle is counted. The detected square area is excluded from the further search and the search algorithm that looks for the first bright pixels continues.

After the detection is completed, information on all detected particles is summarized and stored in the database for further processing.

As part of the detection of dust particle sizes, the found dust particles are sorted by size and a size histogram is created (Figure 5), or a graph of the frequency of occurrence of particles of a certain size in a given detection time.

The advantage of the described invention is the possibility of detecting the size of dust particles according to their actual size, and not just information about the size interval in which the particle is located. Distinguishing

- 3 CZ 309660 B6 the ability of detection is given by the resolution of the digital camera that captures the illuminated dust particles.

Clarification of drawings

The invention is explained in more detail with the help of the drawings in which:

in figure 1 is a device for measuring the size of dust particles in space and time, in figure 2 is a digitized image and its interpretation in the form of a matrix of values, in figure 3 is a binarized image and its interpretation in the form of a matrix of values, in figure 4 is the procedure for finding the dust particles in the image, Figure 5 is a histogram of the sizes of dust particles in the image, Figure 6 is a diagram of a device for measuring the distribution of dust particles in space and time.

An example of the implementation of the invention

According to Figure 6, the device for measuring the distribution of dust particles in space and time consists of an opening 1 for the entry of a gaseous medium containing dust particles 6, a circulation fan 2, a tunnel for directing the medium, a detection chamber with a laser beam 3, a digital camera 4 with optics, an evaluation device 5 on the basis of a personal computer and databases 7 of evaluated data, or the size of dust particles and the frequency of their occurrence over time. The individual components of the device are arranged one behind the other in a tube or any other suitable box, while the digital camera 4 is bidirectionally connected to the evaluation device 5.

Algorithms for background correction for systematic errors, an algorithm for thresholding and binarization of the image matrix form, an algorithm for searching the image matrix form and finding areas corresponding to illuminated dust particles, an algorithm for calculating the size of the found areas and an algorithm for converting the calculated values into database.

Using the circulating fan 2, the measured medium is forced into the device through the opening 1. The dust particles 6 contained in the medium are illuminated by a laser 3 and their image is captured by means of a digital camera 4. The digital image is converted into a matrix form, where each pixel corresponds to one particular cell of the matrix. The digital image has a resolution of 1920 x 1080, or the matrix size corresponds to 1920 in the X direction and 1080 in the Y direction.

Subsequently, the image in the form of a matrix of cells is transferred to the evaluation device 5. In the evaluation device 5, a data processing unit with a CPU is present, which initiates a set of algorithms.

In the evaluation device 5, background correction for systematic errors is performed in the digital image represented by a two-dimensional matrix. Furthermore, a binarized image containing only black and white pixels is created in the evaluation device 5, or the values of the cells of the two-dimensional matrix are overwritten so that they take only one of the two possible values. This operation is performed using the matrix image binarization algorithm. In the evaluation device 5, the algorithm for searching the matrix form of the image is then initiated and the areas corresponding to the illuminated dust particles are searched for. Furthermore, the algorithm for calculating the size of the found areas is initiated in the evaluation device 5, and finally the algorithm for transferring the calculated values to the database is initiated. The data stored in the database are transferred to external processing and based on them, it is then possible to define the air quality in a given place and time.

- 4 CZ 309660 B6

Industrial applicability

The invention can be used to detect the size of dust particles in a gaseous medium in a given space and time. It is possible to continuously measure the content of particles in the air in the internal environment of buildings, in the areas of industrial enterprises, laboratories with controlled gas medium quality, etc., and thereby determine the degree of health risk in a given location.

Claims (2)

1. The method of measuring the distribution of dust particles in space and time, characterized by the fact that the following steps are carried out sequentially: a) a gaseous medium is forced into the measured space, b) the space is illuminated with a laser beam, c) a digital image of laser-illuminated dust particles is created, d) the digital image is converted into matrix form, e) background correction for systematic errors is carried out, f) thresholding and binarization of the matrix form of the image is performed, g) the matrix form of the image is searched and the areas corresponding to the illuminated dust particles are searched for, h) the size of the found areas is calculated a i) the calculated values are transferred to the database of dust particle sizes. 2. Device for measuring the distribution of dust particles in space and time according to claim 1, characterized in that it consists of a sequentially arranged circulation fan (2), a laser (3) and a digital camera (4) with optics, an evaluation device ( 5), where algorithms are integrated in the evaluation device (5) for background correction for systematic errors, for thresholding and binarization of the image in matrix form, for searching the matrix form of the image, for searching for areas corresponding to illuminated dust particles and for conversion to a database of dust particle sizes , while the digital camera (4) is bidirectionally connected to the evaluation device (5).