CN113177911A

CN113177911A - Method for nondestructive evaluation of ozone sensitivity of plants by leaves

Info

Publication number: CN113177911A
Application number: CN202110391446.1A
Authority: CN
Inventors: 张巍巍; 方得安; 许佳瑶; 孙子程
Original assignee: Shenyang University
Current assignee: Shenyang University
Priority date: 2021-04-13
Filing date: 2021-04-13
Publication date: 2021-07-27

Abstract

The invention relates to a method for nondestructive evaluation of ozone sensitivity of plants by leaves, belongs to the field of biological monitoring of atmospheric pollution, and can be applied to nondestructive and continuous measurement of damage symptoms of ozone-sensitive plants such as soybean, poplar, white birch, tilia amurensis, quercus liaotungensis and the like. The method comprises the following steps: classifying damage types according to the leaf surface images, then segmenting the leaf surface images, extracting and analyzing characteristics such as symptom shapes, area sizes and the like, establishing an extraction method based on an ISODATA algorithm under a Lab color mode, and grading damage degrees after obtaining the lesion rate (figure 1). Compared with the existing visual observation or destructive sampling method, the method disclosed by the invention can analyze the damaged area and proportion of the leaf surface without damage, realize continuous monitoring and accurate grading, solve the problem that the damage degree cannot be quantitatively compared due to the difference of ozone damage symptoms of different plants, simply and directly judge the ozone damage of the plants, and has the advantages of rapidness, no damage, continuous observation and the like.

Description

Method for nondestructive evaluation of ozone sensitivity of plants by leaves

Technical Field

The invention relates to a method for nondestructive evaluation of plant ozone sensitivity by leaves, belongs to the field of air pollution biological monitoring, and is particularly suitable for nondestructive, continuous monitoring and quantitative evaluation of plant ozone damage under natural conditions.

Background

The high-concentration atmospheric ozone can generate oxidation damage to plants, can cause the metabolism of the plants to be changed in different degrees, causes the physiological and biochemical processes such as plant pigment content, tissue structure, cell ultrastructure and the like to be changed, and causes the plants to present external damage symptoms. The external symptoms are reflected on the image and show differences of color, shape and the like, and a basis is provided for evaluating the ozone sensitivity of the plants by utilizing an image processing technology. The ozone damage to the plants is not detected as any residual element, but can cause the upper leaves in the canopy to have the symptoms of spots, plaques, dry-out and the like. Moreover, the damage degree of plants is usually proportional to the superficial damage area of the leaf surface, and the color and shape of the lesion are changed and the area is increased as the ozone concentration is increased. Therefore, under natural conditions, the influence degree of the ozone on the plants can be judged and evaluated according to the change of the foliar injury symptoms.

At present, the evaluation method of plant ozone damage mainly comprises the steps of observing or cutting off the scanning area of leaves by naked eyes, then roughly estimating the damage proportion and simply grading. The method has strong subjectivity, cannot carry out rapid and nondestructive quantitative detection, cannot realize continuous monitoring of the development condition of the injury symptom of the same leaf after the leaves are reduced, also has certain influence on plant growth, is time-consuming and labor-consuming, and is not suitable for being developed in a large range. The invention solves the problems of strong subjectivity, incapability of quantifying, time and labor consumption, influence on plant growth and the like of the plant ozone injury symptom analysis method. The method can analyze the damaged area and proportion of the leaf surface without damage, realize continuous monitoring and accurate grading, solve the problem that the damage degree cannot be quantitatively compared due to the difference of ozone damage symptoms of different plants, simply and directly judge the ozone damage of the plants, and has the advantages of rapidness, no damage, continuous observation and the like.

Disclosure of Invention

(1) Data acquisition: under natural conditions, a camera or a mobile phone is used for continuously collecting color image data (RGB, JPG) of the leaf surface of the ozone-sensitive plant every 7-20 days, and the target leaf surface is required to occupy most of the image.

(2) Image processing: classifying the damage types of the leaf surfaces, extracting and analyzing the shape and area characteristics by adopting image processing software by segmenting the damage images of the leaf surfaces, completing the separation of the normal parts of the leaf surfaces from the lesion areas, and establishing an extraction method based on an ISODATA algorithm under a Lab color mode.

(3) Quantitative analysis of injury: for the leaves of the same leaf or the corresponding position of the same plant, the damage area and proportion are calculated by calculating the difference value between the normal leaf image and the whole leaf image, and the damage degree of the leaves is quantitatively analyzed (figure 1). The specific operation steps are as follows.

Step 1: collection plant leaf image (RGB, JPG)

And acquiring a blade color image under a natural condition by using a color digital camera or a mobile phone camera. It is desirable to capture the image in as simple a background as possible so that the target leaf for analysis can occupy a large area of the image.

Step 2: segmenting the blade image, removing the background, and extracting the target blade from the complex background;

and 3, extracting the blades shot in a natural state from a complex background by using drawing 3D software built in a win10 system. The specific operation is as follows: firstly, determining a target blade in an image, detecting a blade boundary by using a painting tool (magic painting brush), and obtaining a complete target blade image with a complex background removed by adjusting, adding and deleting complex backgrounds such as soil, branches and sky, so that the gray level of the background is 0, and the target blade presents an original color image.

And 3, step 3: symptom classification

Ozone does not generally affect the veins, and the symptoms of injury usually occur in the mesophyll portion between the veins, with old leaves more severe than new ones. The leaves with obvious insect spots, protruding mildew spots and the like are removed, the color characteristics are used as the basis for distinguishing, and the ozone-damaged leaves are divided into two categories: blotchy and ulcerated leaves (fig. 2). The invention mainly takes the two types of blades as analysis objects:

a. leaf blotches: fine spots are scattered on the leaf tips and the leaf edges and are white, yellow brown or tan; sometimes full leaf spots appear;

b. ulcer leaves: yellowish brown or brownish plaques with irregular shapes and different sizes appear in the gaps of veins; the color of the lesion becomes dark and dry, and the tip of the leaf burns, curls and falls off when the lesion is serious.

And 4, step 4: performing image segmentation to distinguish leaf region and lesion region

The leaf scab characteristics (color, shape and texture characteristics) are extracted, and the complete target leaf color picture is converted into a gray picture to obtain a corresponding gray image only containing black, white and different shades of gray. Then Image J software is utilized, an ISODATA algorithm is adopted to quickly find an optimal threshold value, the separation of the normal part of the leaf from the lesion area is completed, and an ozone injury symptom extraction method based on an Lab color mode and adopting an ISODATA algorithm (iterative self-organizing data analysis method) is established;

image J software analysis step:

①Open

opening the target leaf;

②Image/Type/Lab Stack/

converting the color image from the RGB model into an L × a × b image, and converting the L × or a × component into an 8-bit image;

③Image/Adjust/Threshold/Default

and processing the histogram of the L or a component by using a threshold segmentation method of an ISODATA algorithm to obtain an optimal threshold image, wherein the area of the optimal threshold image is the area of the green leaf after the lesion is removed. Filling the whole blade to obtain an image which is the maximum area of the blade;

④Process/Binary/Make binary/

and (4) carrying out binarization processing on the optimal threshold value image obtained by segmentation to enable the image to show obvious black-white differentiation, and obtaining an image of a normal part of the leaf. After binarization processing is carried out on the maximum leaf image, if a hole is found in the leaf, filling the leaf by Fill Holes to obtain the whole leaf image;

⑤Process/Image Calculator

because the ozone often causes the plaque at the edge of the blade, the area of the marginal plaque can be underestimated by directly reversing the image of the normal part of the blade after binarization, so that an accurate plaque image is obtained by adopting a method of subtracting the normal part image of the blade from the whole blade image;

⑥Analyze/Analyze Particles/

in this way, the area of each lesion, the total area of the lesion and the maximum leaf area can be calculated.

And 5, step 5: quantitative analysis of ozone damage degree of blade

The size and proportion of the damaged area are indexes for measuring the damage degree of the blade caused by the ozone. According to the formula: the lesion rate (%) = total lesion area/total leaf area x 100, and the lesion rate was calculated. Can carry out quick and nondestructive quantitative detection on the lesion spots of a large number of leaves.

Drawings

FIG. 1 is a method and a flow chart for nondestructive evaluation of ozone sensitivity of plants by leaves.

Figure 2 two typical ozone injury symptoms.

FIG. 3 is a color picture of ozone foliar damage of Quercus mongolicus.

FIG. 4 is a graph of the target foliar damage of Quercus mongolicus.

FIG. 5 is an ozone damage area analysis diagram of a quercus mongolica target blade.

Fig. 6 is a color picture of ozone leaf surface damage of birch.

Fig. 7 graph of birch target leaf surface damage.

Fig. 8 is a normal area diagram of a birch target leaf.

Fig. 9 is a graph of an area analysis of a target leaf damage of birch.

Detailed Description

The invention will be further illustrated with reference to the following specific examples.

Example analysis 1

In this example, quercus mongolica was selected as the subject, and foliar damage analysis and ozone sensitivity evaluation were performed by the method of the present invention:

step 1: collecting plant leaf images (RGB, JPG) to expand target leaves as much as possible (FIG. 3);

step 2: using a win10 system to make the background black by a magic painting pen function with drawing 3D software, obtaining a target leaf and extracting the target leaf from the complex background (figure 4);

and 3, step 3: classifying the symptoms according to the color and texture characteristics, and judging as follows: leaf of macula;

and 4, step 4: and (4) carrying out Image segmentation, distinguishing a leaf area and a lesion area, and carrying out Image J operation steps of (i) - (sixth), so as to obtain the lesion area and the whole leaf area. First, open the target leaf file, convert the color image from the RGB model to L x a b image, convert the a component to 8-bit image. And operating Image/Adjust/Threshold/Default to obtain the normal area Image of the leaf. And converting the a component into an 8-bit Image again, operating Image/Adjust/Threshold/Default, filling the whole blade, clicking the appliance, and operating Process/Binary/Fill Holes to obtain the whole blade Image. And operating the Process/Image Calculator to obtain an accurate Image of the lesion area. Operating Analyze/Analyze partitions to calculate the area of each lesion, the total area of the lesion, and the maximum leaf area (FIG. 5);

and 5, step 5: and (5) quantitatively analyzing the ozone damage degree of the blade. According to the formula: the lesion ratio (%) was calculated as = total lesion area/total leaf area × 100. The damage ratio (% spotting) is = 46150 ÷ 3215182 × 100, 1.44.

Example analysis 2

In this example, white birch was selected as the target, and the method of the present invention was used to perform foliar damage analysis and ozone sensitivity evaluation:

step 1: collecting plant leaf images (RGB, JPG) to expand target leaves as much as possible (FIG. 6);

step 2: using a win10 system to make the background black by a magic painting pen function with drawing 3D software, obtaining a target leaf and extracting the target leaf from the complex background (figure 7);

and 3, step 3: classifying the symptoms according to the color and texture characteristics, and judging as follows: ulcerated leaves;

and 4, step 4: and (4) carrying out Image segmentation, distinguishing a leaf area and a lesion area, and carrying out Image J operation steps of (i) - (sixth), so as to obtain the lesion area and the whole leaf area. First, open the target leaf file, convert the color image from the RGB model to L x a b image, convert the L x component to 8-bit image. Operating Image/Adjust/Threshold/Default, obtaining the normal area Image of the leaf (figure 8). And converting the a component into an 8-bit Image again, operating Image/Adjust/Threshold/Default, filling the whole blade, clicking the appliance, and operating Process/Binary/Fill Holes to obtain the whole blade Image. And operating the Process/Image Calculator to obtain an accurate Image of the lesion area. Operating Analyze/Analyze partitions to calculate the area of each lesion, the total area of the lesion, and the maximum leaf area (FIG. 9);

and 5, step 5: and (5) quantitatively analyzing the ozone damage degree of the blade. According to the formula: the lesion ratio (%) was calculated as = total lesion area/total leaf area × 100. The damage ratio (% spotting) is = 471836 ÷ 2987415 × 100, 15.8.

The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be able to cover the technical solutions and the inventive concepts of the present invention within the technical scope of the present invention.

Claims

1. A method for evaluating the ozone sensitivity of plants without damaging leaves is characterized by comprising the following steps: (1) data acquisition: under natural conditions, using a camera or a mobile phone to continuously collect color image data (RGB, JPG) of the leaf surface of the ozone-sensitive plant, wherein the target leaf surface is required to occupy most of the image; (2) image processing: classifying the damage types of the leaf surfaces, segmenting the damage images of the leaf surfaces, extracting and analyzing the shape and area characteristics by adopting image processing software, completing the separation of the normal parts of the leaf surfaces from the lesion areas, and establishing an extraction method (3) quantitative analysis of the damage based on an ISODATA algorithm under a Lab color mode: and calculating the difference value between the normal leaf image and the whole leaf image of the same leaf to the leaf of the same leaf or the corresponding position of the same plant, calculating the damage area and proportion, and quantitatively analyzing the damage degree of the leaf.

2. The method for the nondestructive evaluation of ozone sensitivity of plants according to claim 1, wherein the ozone damage symptoms include spotted and ulcerated damage types.

3. The method for the nondestructive evaluation of the ozone sensitivity of the plants on the leaves as claimed in claim 1, wherein the ozone damage symptom extraction method based on the ISODATA algorithm under the Lab color mode (L or a component) is established for the leaf damage image.

4. The method for the nondestructive evaluation of ozone sensitivity of plants on leaves as claimed in claim 1, wherein the quantitative analysis method for the damage of leaves comprises calculating the difference between the normal leaf image and the whole leaf image of the leaves at the corresponding positions of the same leaf or the same plant, calculating the damage area and ratio, and quantitatively analyzing the damage degree of the leaves.