阅读说明：本技术 一种基于图像识别的球团固结程度评价方法 (Pellet consolidation degree evaluation method based on image recognition ) 是由 王耀祖 贺威 刘征建 张建良 于欣波 侯静怡 马黎明 马云飞 于 2021-07-12 设计创作，主要内容包括：本发明公开了一种基于图像识别的球团固结程度评价方法,包括：制备球团；采用扫描电子显微镜或者矿相显微镜获取球团图像；对获取的图像进行图像识别,得到图像中颗粒的数量、周长和面积；基于图像识别结果提出球团固结评价体系,计算出球团内部颗粒生长指数、颗粒均匀性指数、颗粒固结指数及焙烧熟化度。本发明基于球团矿矿相结构图像,采用分水岭分割、卷积神经网络等智能算法来实现球团内部颗粒的分割和识别,并自动量化获取球团内部颗粒的数量、颗粒边界周长、颗粒面积等参数,得到球团矿内部颗粒生长指数、颗粒均匀性指数、颗粒固结指数及颗粒熟化度,对评价铁矿粉的连晶性能以及优化球团矿的焙烧工艺参数具有重要的意义。(The invention discloses an image recognition-based pellet consolidation degree evaluation method, which comprises the following steps of: preparing pellets; acquiring a pellet image by adopting a scanning electron microscope or a mineral phase microscope; carrying out image recognition on the obtained image to obtain the number, perimeter and area of particles in the image; and providing a pellet consolidation evaluation system based on the image recognition result, and calculating the grain growth index, the grain uniformity index, the grain consolidation index and the roasting curing degree in the pellet. According to the invention, based on the pellet phase structure image, the segmentation and identification of pellets inside the pellet are realized by adopting intelligent algorithms such as watershed segmentation, convolutional neural network and the like, and the parameters such as the number of the pellets inside the pellet, the perimeter of the particle boundary, the particle area and the like are automatically and quantitatively obtained, so that the growth index, the uniformity index, the consolidation index and the particle curing degree of the pellets inside the pellet are obtained, and the method has important significance for evaluating the continuous crystallization performance of iron ore powder and optimizing the roasting process parameters of the pellet.)

1. The pellet consolidation degree evaluation method based on image recognition is characterized by comprising the following steps of:

step one, preparing pellets;

acquiring a pellet image by adopting a scanning electron microscope or a mineral phase microscope, and characterizing the microstructure of the pellet;

step three, carrying out image recognition on the obtained image to obtain the number, the perimeter and the area of particles in the image;

and step four, providing a pellet consolidation evaluation system based on the image recognition result, and calculating the grain growth index, the grain uniformity index, the grain consolidation index and the roasting curing degree in the pellet.

2. The method for evaluating the degree of pellet consolidation based on image recognition as claimed in claim 1, wherein the process of preparing pellets in the first step comprises:

fully drying a predetermined amount of iron-containing raw material at 105 ℃ to remove bulk water; taking 50g of iron-containing raw material and 4ml of deionized water to be fully mixed, and then taking 15g of the fully mixed raw material to be pressed under the pressure of 10MPa to prepare columnar pellets with the height of 10mm and the diameter of 20 mm;

drying the obtained pellets in a drying box at 105 ℃ for 3h to remove free water in the pellets;

placing the dried pellets in an alumina porcelain boat for oxidizing roasting, preheating for 5min under the condition of Ar protective atmosphere with the flow rate of 3L/min, converting gas into compressed air with the flow rate of 5L/min when the temperature of a sample reaches 850 ℃, and oxidizing for 30 min;

after the oxidation is finished, transferring the pellets to another high-temperature furnace for high-temperature roasting for 30min, and introducing air with the flow rate of 5L/min in the whole roasting process;

after the pellet roasting is finished, cooling the pellet at ambient temperature; in the experimental process, the pellet which is not roasted is used as a blank control group of the experiment, each group of experiment is repeated for three times, and the average value of the experimental results is taken;

after the experiment is finished, the sample is embedded in epoxy resin, and water is used as a lubricant to carry out grinding and polishing treatment on the sample.

3. The method for evaluating the degree of pellet consolidation based on image recognition of claim 2, wherein the iron-containing raw material refers to iron-containing materials capable of being used for agglomeration, and comprises iron ore powder, steelmaking sludge and iron-containing dust.

4. The method for evaluating the degree of pellet consolidation based on image recognition as claimed in claim 2, wherein the oxidizing roasting schedule can be replaced by any schedule suitable for densification of powder materials.

5. The method for evaluating the degree of pellet consolidation based on image recognition as claimed in claim 1, wherein in the second step, all the images are obtained under the same magnification, and the consolidation index data of each sample is obtained from at least 10 pictures of the fracture surface pellet phase structure.

6. The method for evaluating the degree of pellet consolidation based on image recognition of claim 1, wherein in the third step, the image recognition of the acquired image specifically comprises:

carrying out gray level conversion, median filtering and expansion mask processing on the obtained image;

dividing particles in the image by a watershed algorithm, and performing intelligent calculation by using a convolutional neural network;

the number, perimeter and area of particles in the output image.

7. The method for evaluating the degree of pellet consolidation based on image recognition of claim 1, wherein in the fourth step, the expression of the particle consolidation index is as follows:

wherein CI is the pellet consolidation index, SPP_{oxidized pellet}The total circumference of the pellets after preheating, SPP_{roasted pellet}Is the total perimeter of the particles inside the pellet after roasting.

8. The method for evaluating the degree of pellet consolidation based on image recognition as claimed in claim 1, wherein in the fourth step, the expression of the grain growth index is as follows:

wherein GI is the grain growth index, APA, of the pellets after roasting_{roasted pellet}Is the area of the oxidized pellet particles, APA_{oxidized pellet}Is the average area of the pellet particles.

9. The method for evaluating the degree of pellet consolidation based on image recognition as claimed in claim 1, wherein in the fourth step, the expression of the particle uniformity index is as follows:

wherein UI is a particle uniformity index showing the degree of uniformity of particle size within the pellet, A_iIs the area of the ith particle,is the average area of the particles and N is the total number of particles within the pellet.

10. The method for evaluating the degree of pellet consolidation based on image recognition of claim 1, wherein in the fourth step, the expression of roasting maturity is as follows:

wherein RI is a baking aging degree,is greater than 4500 μm²Area of the particles of (A)_TotalIs the total area of the particles inside the pellet.

Technical Field

The invention relates to the technical field of iron ore oxidized pellets, in particular to a pellet consolidation degree evaluation method based on image recognition.

Background

The pellet is used as an important iron-containing raw material for blast furnace ironmaking, and is a sphere with the diameter of 8-16 mm. The pellet preparation comprises the processes of pelletizing and preheating roasting, and comprises the mixing of raw materials such as fine-grained ore, dolomite, bentonite and the like, and then the oxidation and roasting are carried out under the high-temperature condition to achieve chemical components, mechanical properties and metallurgical properties required by proper smelting. Wherein, the technical parameters such as roasting temperature, roasting time and the like have great influence on the mineral composition, mechanical strength and metallurgical performance of the pellet. It is well known that the consolidation mechanism of pellets relies mainly on solid phase reactions, including diffusion between single component particles to form bridges (sintering necks), and diffusion between multi-component systems to form compounds and solid solutions. The consolidation temperature mainly occurs below the melting point temperature of the oxide, and a small amount of liquid phase is generated in the roasting process. In the process of producing the pellet, due to the diversity of iron ores, in order to optimize the performance of the pellet as much as possible, technical parameters of a roasting process are needed to be adjusted to optimize the internal grain characteristics of the pellet, including the shape of a grain boundary, the size of a grain, the shape of pores and the like, so that the overall optimization of the metallurgical performance and the mechanical performance of the pellet is realized. Therefore, the unification of the microscopic property and the macroscopic property of the pellet has important significance for improving the metallurgical property of the pellet and evaluating the roasting property of the magnetite powder.

According to literature research, there are few references to quantitative characterization of the degree of sintering of pellets, and H O,Bailon A M G,de Alves F J,et al.Methodology development for determining the sintering degree in iron ore pellet grains through automatic image analysis[C]// Proceedings of the 6th International consistency on the Science and Technology of ironmaking-ICSTI, 2012,14.) A method for measuring the degree of sintering of pellets based on optical microscopy was proposed, which they believe enabled quantitative determination of the size of the crystal grains and the sintering necks between the grains. Literature (Newros F, thumb M J, Jonsson H, et al]Pattern Recognition, Elsevier,2015,48(11): 3451-3465) based on the connection structure and binding characteristics of the particles, a sintering degree quantification processing method based on image Recognition is proposed, and the radius, curvature and density of the particles in the pellet under four different sintering temperature conditions are calculated. Literature (Kumart K S, Simonsson M, Viswanathan N, et al, assessing a novel method to correct the macromolecular and microscopic degree of morphological in magnetic pellet reduction indication J]Steel research international,2018,89(3): 1700366) the degree of pellet firing was determined using distance conversion quantification based on image processing methods.

All the researches provide a good research idea for quantitative analysis of the roasting degree of the pellet. In previous studies, however, the scholars characterized the size of the sintering necks primarily by the diameter and size of the grains. However, due to the small size of the particles inside the pellet, the irregularity of the shape, and the limitation of the resolution of the picture, this statistical method tends to cause relatively large errors in the characterization of the sintering neck and the particle size. More importantly, the growth behavior of the particles, such as the growth index of the mineral powder during roasting, the uniformity index of the crystal grains, the temperature sensitivity during continuous crystallization, and the like, does not form a uniform high-temperature evaluation index.

Aiming at the technical problems, the invention aims to provide a pellet consolidation degree evaluation method based on image recognition, which can realize evaluation of particle characteristic parameters and consolidation degree in a pellet based on the image recognition technology, systematically analyze high-temperature performances such as pellet consolidation index, particle growth index and particle uniformity index and comprehensively reflect the relation between the microstructure and the macroscopic performance of the pellet. The method is simple and quick to operate, takes factors into consideration systematically and comprehensively, enriches the image recognition evaluation index of the pellet consolidation degree, reduces the limitation of the image recognition technology caused by resolution, and provides detailed standards for evaluating the pellet roasting degree later, representing the growth behavior of sintering necks, the grain crystal connection degree and the like; has certain guiding significance for optimizing the pellet roasting process and improving the pellet strength.

Disclosure of Invention

The invention aims to provide a pellet consolidation degree evaluation method based on image recognition, which is based on a pellet ore phase structure image, adopts intelligent algorithms such as watershed segmentation, convolutional neural network and the like to realize the segmentation and recognition of pellets inside the pellet, automatically quantifies and obtains parameters such as the number of pellets inside the pellet, the perimeter of a particle boundary, the area of the particles, the curvature of the particles and the like, and obtains a pellet growth index, a particle uniformity index, a particle consolidation index and a particle curing degree inside the pellet. The method can accurately describe the growth and consolidation degree of the ore powder particles in the roasting process of the pellet based on the image recognition technology, and has very important significance for evaluating the continuous crystallization performance of the iron ore powder and optimizing the roasting process parameters of the pellet.

To solve the above technical problem, an embodiment of the present invention provides the following solutions:

a pellet consolidation degree evaluation method based on image recognition comprises the following steps:

step one, preparing pellets;

acquiring a pellet image by adopting a scanning electron microscope or a mineral phase microscope, and characterizing the microstructure of the pellet;

step three, carrying out image recognition on the obtained image to obtain the number, the perimeter and the area of particles in the image;

and step four, providing a pellet consolidation evaluation system based on the image recognition result, and calculating the grain growth index, the grain uniformity index, the grain consolidation index and the roasting curing degree in the pellet.

Preferably, the process for preparing the pellets in the first step comprises the following steps:

fully drying a predetermined amount of iron-containing raw material at 105 ℃ to remove bulk water; taking 50g of iron-containing raw material and 4ml of deionized water to be fully mixed, and then taking 15g of the fully mixed raw material to be pressed under the pressure of 10MPa to prepare columnar pellets with the height of 10mm and the diameter of 20 mm;

drying the obtained pellets in a drying box at 105 ℃ for 3h to remove free water in the pellets;

placing the dried pellets in an alumina porcelain boat for oxidizing roasting, preheating for 5min under the condition of Ar protective atmosphere with the flow rate of 3L/min, converting gas into compressed air with the flow rate of 5L/min when the temperature of a sample reaches 850 ℃, and oxidizing for 30 min;

after the oxidation is finished, transferring the pellets to another high-temperature furnace for high-temperature roasting for 30min, and introducing air with the flow rate of 5L/min in the whole roasting process;

after the pellet roasting is finished, cooling the pellet at ambient temperature; in the experimental process, the pellet which is not roasted is used as a blank control group of the experiment, each group of experiment is repeated for three times, and the average value of the experimental results is taken;

after the experiment is finished, the sample is embedded in epoxy resin, and water is used as a lubricant to carry out grinding and polishing treatment on the sample.

Preferably, the iron-containing raw material refers to iron-containing materials capable of being used for agglomeration, and comprises iron ore powder, steelmaking sludge and iron-containing dust.

Preferably, the oxidative calcination regime can be replaced by any regime suitable for densification of powdered materials.

Preferably, in the second step, all the images are acquired under the same magnification, and each sample consolidation index data is acquired from at least 10 pictures of the cross-section pellet phase structure.

Preferably, in the third step, the image recognition of the acquired image specifically includes:

carrying out gray level conversion, median filtering and expansion mask processing on the obtained image;

dividing particles in the image by a watershed algorithm, and performing intelligent calculation by using a convolutional neural network;

the number, perimeter and area of particles in the output image.

Preferably, in the fourth step, the particle consolidation index is expressed as:

wherein CI is the pellet consolidation index, SPP_{oxidized pellet}The total circumference of the pellets after preheating, SPP_{roasted pellet}Is the total perimeter of the particles inside the pellet after roasting.

Preferably, in the fourth step, the particle growth index expression is as follows:

wherein GI is the grain growth index, APA, of the pellets after roasting_{roasted pellet}Is the area of the oxidized pellet particles, APA_{oxidized pellet}Is the average area of the pellet particles.

Preferably, in the fourth step, the particle uniformity index expression is as follows:

wherein UI is a particle uniformity index showing the degree of uniformity of particle size within the pellet, A_iIs the area of the ith particle,is the average area of the particles and N is the total number of particles within the pellet.

Preferably, in the fourth step, the roasting curing degree expression is as follows:

wherein RI is a baking aging degree,is greater than 4500 μm²Area of the particles of (A)_TotalIs the total area of the particles inside the pellet.

The technical scheme provided by the embodiment of the invention has the beneficial effects that at least:

the embodiment of the invention provides a pellet consolidation degree evaluation method based on image recognition, which can be used for representing the roasting consolidation degree of pellet ore or other porous materials on a particle scale. The method has the following advantages: firstly, the method has comprehensive evaluation indexes, provides four standards for evaluating the pellet consolidation degree, namely a particle Consolidation Index (CI), a particle Growth Index (GI), a particle Uniformity Index (UI) and a roasting curing degree (RI), and obviously improves the scientificity of representing pellet consolidation by microscopic morphology; secondly, the method adopts intelligent algorithms such as deep learning to realize the segmentation and identification of the particles inside the pellet, and automatically quantifies and obtains parameters such as the number of the particles inside the pellet, the perimeter of the particle boundary, the particle area, the particle curvature and the like, thereby greatly reducing errors caused by low image identification resolution, poor contrast and the like; in addition, the method applies an intelligent algorithm to divide and identify the particles in the pellet, avoids the error of manual operation, ensures the precision of evaluation indexes, and has important significance for the evaluation of the roasting consolidation degree of porous materials such as the pellet and the like.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a flowchart of a pellet consolidation degree evaluation method based on image recognition according to an embodiment of the present invention;

FIGS. 2 a-2 f are schematic diagrams of SEM images of pellets according to the present invention illustrating the processing of weak particle connection and separation by intelligent segmentation;

fig. 3 a-3 l are graphs showing the results of the SEM images of the pellets in the example of the present invention using image recognition processing.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The embodiment of the invention provides a pellet consolidation degree evaluation method based on image recognition, and as shown in fig. 1, the method comprises the following steps:

the pellet preparation method comprises the following specific steps:

fully drying a predetermined amount of iron-containing raw material at 105 ℃ to remove bulk water; taking 50g of iron-containing raw material and 4ml of deionized water to be fully mixed, and then taking 15g of the fully mixed raw material to be pressed under the pressure of 10MPa to prepare columnar pellets with the height of 10mm and the diameter of 20 mm;

drying the obtained pellets in a drying box at 105 ℃ for 3h to remove free water in the pellets;

placing the dried pellets in an alumina porcelain boat for oxidizing roasting, preheating for 5min under the condition of Ar protective atmosphere with the flow rate of 3L/min, converting gas into compressed air with the flow rate of 5L/min when the temperature of a sample reaches 850 ℃, and oxidizing for 30 min;

after the oxidation is finished, transferring the pellets to another high-temperature furnace for high-temperature roasting for 30min, and introducing air with the flow rate of 5L/min in the whole roasting process;

after the pellet roasting is finished, cooling the pellet at ambient temperature; in the experimental process, the pellet which is not roasted is used as a blank control group of the experiment, each group of experiment is repeated for three times, and the average value of the experimental results is taken;

after the experiment is finished, the sample is embedded in epoxy resin, and water is used as a lubricant to carry out grinding and polishing treatment on the sample.

Wherein the iron-containing raw material refers to all iron-containing materials for agglomeration, and can be iron ore powder, steelmaking sludge, iron-containing dust and the like. The oxidizing roasting system can be any system suitable for the densification of the powder material and is not limited to the oxidizing roasting of the pellet.

In addition, it should be noted that the experimental parameters are all preferred parameters, and do not limit the present invention, that is, the embodiment of the present invention does not strictly limit the agglomeration process, equipment, and ingredient ratio, and is suitable for all types of agglomeration or pellets.

And step two, acquiring a pellet image by adopting a scanning electron microscope or a mineral phase microscope, and characterizing the microstructure of the pellet.

In the step, all the images are obtained under the same magnification, and the consolidation index data of each sample are obtained from the pictures of pellet phase structures of at least 10 sections.

And step three, carrying out image recognition on the obtained image to obtain the number, the perimeter and the area of the particles in the image.

In this step, the image recognition of the acquired image includes the steps of:

carrying out gray level conversion, median filtering and expansion mask processing on the obtained image;

dividing particles in the image by a watershed algorithm, and performing intelligent calculation by using a convolutional neural network;

the number, perimeter and area of particles in the output image.

The image processing in the embodiment of the invention adopts intelligent algorithm recognition, realizes the representation of the whole structure of the pellet by controlling the early-stage sample and the experimental magnification, and is suitable for evaluating the consolidation degree of the pellet, the briquetting or the porous material.

And step four, providing a pellet consolidation evaluation system based on the image recognition result, and calculating a pellet internal particle Growth Index (GI), a particle Uniformity Index (UI), a particle Consolidation Index (CI) and a roasting curing degree (RI).

Specifically, the particle consolidation index expression is:

wherein CI is the pellet consolidation index, SPP_{oxidized pellet}The total circumference of the pellets after preheating, SPP_{roasted pellet}Is the total perimeter of the particles inside the pellet after roasting.

The grain growth index expression is:

wherein GI is the grain growth index, APA, of the pellets after roasting_{roasted pellet}Is the area of the oxidized pellet particles, APA_{oxidized pellet}Is the average area of the pellet particles.

The particle uniformity index expression is:

wherein UI is a particle uniformity index showing the degree of uniformity of particle size within the pellet, A_iIs the ith granuleThe area of the particles is such that,is the average area of the particles and N is the total number of particles within the pellet.

The roasting curing degree expression is as follows:

wherein RI is a baking aging degree,is greater than 4500 μm²Area of the particles of (A)_TotalIs the total area of the particles inside the pellet.

The process of the invention is illustrated in detail below by means of two specific examples.

The first embodiment is as follows: and (3) evaluating the roasting consolidation degree of the high-iron low-silicon magnetite pellets with the TFe content of 71.66% and the SiO2 of 0.19%.

50g of high-iron low-silicon magnetite ore powder is taken and fully mixed with 4ml of deionized water, then 15g of iron ore powder is taken and pressed into pellets, and the pressed pellets are dried in a drying oven at 105 ℃ for 3 hours to remove free water in the pellets. Then the pellets were placed in an alumina porcelain boat and preheated for 5min under Ar (3L/min) protective atmosphere. When the temperature reaches 850 ℃, converting the gas into compressed air (5L/min) for oxidation for 30min, after the oxidation is finished, transferring the pellets into another high-temperature furnace for high-temperature roasting for 30min, wherein the roasting temperatures are respectively set to 1200 ℃, 1250 ℃ and 1300 ℃, SEM images of the high-iron low-silicon magnetite pellets can be obtained by analyzing and processing, the areas of particles in the oxidized pellets are mainly concentrated between 0 and 8700 mu m2, the total areas of the particles account for 63.55 percent of the total areas of the pellets, and the porosity is 36.45 percent; when the roasting temperature is 1200 ℃, the particle size inside the pellet is mainly concentrated between 1200 and 5400 mu m, and as the roasting temperature is increased, the particle size is increased, and fine particles basically disappear. It was calculated based on the consolidation evaluation criteria that as the firing temperature increased, the SPP gradually decreased from 37639 μm to 21007 μm, GI increased from 0 to 1.66, UI decreased from 0.43 to 0.16, and CI increased from 0 to 0.44.

Example two: and (3) evaluating the roasting consolidation degree of the low-iron high-silicon magnetite pellets with the TFe content of 64.97% and the SiO2 of 6.90%.

And (3) fully mixing 50g of low-iron high-silicon magnetite ore powder with 4ml of deionized water, then pressing 15g of iron ore powder into pellets, and drying the pressed pellets in a drying box at 105 ℃ for 3 hours to remove free water in the pellets. Then the pellets were placed in an alumina porcelain boat and preheated for 5min under Ar (3L/min) protective atmosphere. When the temperature reaches 850 ℃, converting gas into compressed air (5L/min) for oxidation for 30min, after the oxidation is finished, transferring the pellet ore to another high-temperature furnace for high-temperature roasting for 30min, wherein the roasting temperatures are respectively set to 1200 ℃, 1250 ℃ and 1300 ℃, SEM images of the low-iron high-silicon magnetite pellet ore can be obtained by analyzing and processing, the areas of particles in the oxidized pellet ore are mainly concentrated between 0 and 8700 mu m2, most of the particles are concentrated between 0 and 2100 mu m2, and the total area of the particles accounts for 61.15 percent of the total area of the pellet ore; when the roasting temperature is 1200 ℃, the particle size in the pellets is mainly concentrated between 0 and 13500 mu m, and with the increase of the roasting temperature, the particle size is increased, fine particles basically disappear, and the particle-particle interface is eliminated. It was calculated based on the consolidation level evaluation criteria that as the firing temperature was increased, the SPP gradually decreased from 42049 μm to 25629 μm, the GI increased from 0 to 1.72, the UI decreased from 0.39 to 0.09, and the CI increased from 0 to 0.41.

Specifically, fig. 2 a-2 f are schematic diagrams illustrating SEM images of pellets according to an embodiment of the present invention using smart segmentation to process weak connection and separation of particles. Wherein, FIG. 2a is an original SEM image of pellets under the roasting condition of 1200 ℃; FIG. 2b is the image after the binarization process; FIG. 2c is a programmed image; FIG. 2d is an enlarged SEM image of FIG. 2 a; FIG. 2e is the binarized image; figure 2f is a particle separation and edge identification image. Fig. 3 a-3 l are graphs showing the results of the SEM images of the pellets in the example of the present invention using image recognition processing. Wherein, fig. 3 a-3 c are schematic diagrams of oxidized pellets; FIGS. 3 d-3 f are graphs showing the results of 1200 ℃ treatment; FIGS. 3 g-3 i are graphs showing the results of 1250 deg.C treatment; FIGS. 3 j-3 l) are graphs showing the results of 1300 ℃.

In summary, the invention provides a pellet consolidation degree evaluation method based on image recognition, which is based on a pellet ore phase structure image, and adopts intelligent algorithms such as watershed segmentation and convolutional neural network to realize the segmentation and recognition of pellets inside the pellet, and automatically quantizes and obtains parameters such as the number of pellets inside the pellet, the perimeter of the particle boundary, the particle area, the particle curvature and the like, so as to obtain a pellet inside particle growth index, a particle uniformity index, a particle consolidation index and a particle maturity. The method can accurately describe the growth and consolidation degree of the ore powder particles in the roasting process of the pellet ore based on the image recognition technology, and has very important significance for evaluating the continuous crystallization performance of the iron ore powder and optimizing the roasting process parameters of the pellet ore.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.