KR100737288B1 - Method for searching plants according to outline of leaf image - Google Patents

Abstract

The present invention relates to a method for searching a plant through a leaf-shaped contour, wherein (A) the leaf providing the information providing server computer to correspond to the ecological information, the shape information and the sketch corresponding to each plant and the feature points of the extracted leaf-shaped contour. Storing the shape outline information in each DB; (B) uploading, via the upload unit, leaf shapes such as sketches, photographs, and tables stored in the storage unit through the input unit to search for a specific plant stored in the leaf shape contour information DB of the information providing server computer; (C) the information providing server computer searching through the matching unit for leaf similarity whether the leaf shape similar to the leaf shape uploaded from the user computer is in the leaf shape contour information DB; (D) the information providing server computer displaying a leaf shape similar to the uploaded leaf shape through an output of the user computer through a network; And (E) when the leaf shape similar to the uploaded leaf shape is displayed through the output unit through the step (D), selecting the corresponding leaf shape to display information on each species of the plant having the leaf shape that the user computer wants to search. Identifying.

Sketch, outline, filter, upload, search

Description

Search method of plant through leaf-shaped outline {METHOD FOR SEARCHING PLANTS ACCORDING TO OUTLINE OF LEAF IMAGE}

1 is a view showing the overall system applied to the method of searching for plants through the contour of the leaf-shaped according to the present invention.

Figure 2 is a flow chart showing a method of searching for plants through the contour of the leaf shape according to the present invention.

Figure 3 is a flow chart illustrating in detail the process of extracting the leaf-shaped feature point through the MPP method of step S420 of FIG.

4a and 4b is a view showing the ecological information input form and the input ecological information DB of the plant to be applied to the present invention.

5a and 5b is a view showing a specific example of the form information input form and the input form information DB of the plant to be applied to the present invention.

Figure 6 is a view showing in detail by applying the flow chart of Figure 3 to the leaf shape.

7 is a view showing a connection between points after merging unnecessary points among the extracted feature points of the leaf-shaped contour.

8 is a diagram illustrating an example of extracting feature points of a leaf contour after correcting rotation, movement, and size when the leaf shapes are not symmetrical.

9 illustrates an example of an uploading unit and a display screen of a user computer.

* Explanation of symbols for main parts of drawings *

100: information server computer 110: interface

120: input unit 125: extraction unit

130: matching unit 140: output unit

150: storage unit 151: ecological information DB

152: shape information DB 153: leaf shape contour information DB

200: network 300: user computer

330: filtering unit 355: uploading unit

The present invention relates to a method for searching plants through the contour of leaf shape, and more particularly, by sketching the leaf shape of each plant inhabiting the whole country, and at the same time, the leaf by MPP (Minimum Perimeter Polygon) technique. After extracting the feature points of the shape contour and making it into a database, if a user requests a search through the contours of a specific plant leaf shape, such as a sketch, a photo, and a table, the user searches for a similar plant having the leaf shape requested. And it relates to a method of searching for plants through the contour of the leaf shape to be identified.

In general, plants are organisms that have cell walls and chlorophyll and have photosynthesis, which lead to independent nutrition. Particularly, aquatic plants normally grow in water and come out of the water even when they come out of water. Is a plant that grows in the water, producing substances and circulating nutrients in the water, providing oxygen and food sources and habitats for aquatic organisms such as fish and invertebrates, removing water pollutants and water purification, and maintaining biodiversity of the aquatic ecosystem. It is in charge of many roles, and research on this is being actively conducted.

In addition, in recent years, information on aquatic plants needed for the development of biotope in urban spaces, restoration of wetlands and natural streams, and the creation of ecological pathways along with economic development are needed. In addition, the need for the public to be searchable, so that it can be used as a basic information for research on aquatic plants and various information on each species of aquatic plants.

However, when most users interested in a particular plant search for a plant through a specific site, when the user knows the name of the specific plant, the user searches by entering the name of the plant such as forsythia, buttercup, and lotus as a keyword. However, if you do not know the name of the plant exactly, there is a problem that can not easily identify a variety of information, characteristics and characteristics related to the plant.

The present invention has been made to solve the above problems, the purpose of which is to sketch the leaf shape of each plant inhabiting the whole country and at the same time the contour of the leaf shape by the MPP (Minimum Perimeter Polygon) technique After extracting the feature points and making a database, if the user requests a search through the outline of a specific plant leaf shape, such as a sketch, a photograph, and a table, the user can search for and identify a plant having the leaf shape requested. It is to provide a search method for plants through the outline of a leaf shape.

Plant search method through the contour of the leaf shape according to the present invention for achieving the above object, leaf using a system comprising an information providing server computer 100, network 200 and the user computer 300 As a method for searching each plant through the outline of the shape, (A) the information providing server computer 100 receives the ecological information and the form information corresponding to each plant through the input unit 120 ecological information DB (151) And form information DB (152) (S410), and receives the leaf-shaped sketch of each plant through the input unit 120 and at the same time through the extraction unit 125 of the leaf-shaped contour using the MPP technique Extracting a feature point, and then storing (S420) the correspondence of the sketch with the feature point of the extracted leaf-shaped contour in the leaf-shaped contour information DB 153 (S410 and S420); (B) a sketch previously stored in the storage unit 350 by the user computer 300 through the input unit 320 to search for a specific plant stored in the leaf-shaped outline information DB 153 of the information providing server computer 100, Uploading a leaf shape such as a photograph and a table through the upload unit 355 (S430); (C) Leaf matching similarity of whether leaf information similar to leaf shape uploaded from user computer 300 is in leaf shape contour information DB 153 by the information providing server computer 100 through matching unit 130. Searching (S440); (D) If there is a leaf shape similar to the leaf shape uploaded through the step (C), the information providing server computer 100 through the output unit 340 of the user computer 300 through the network 200; Displaying a leaf shape similar to the uploaded leaf shape (S450); And (E) when the leaf shape similar to the uploaded leaf shape is displayed through the output unit 340 through the step (D), the leaf shape that the user computer 300 searches for is selected by selecting the corresponding leaf shape. It includes the step (S460) of identifying each species information of the plant having, in the step (A), if the feature point of the leaf-shaped contour is irregular, the leaf shape through the extraction unit 125, the movement and size After correcting, it is preferable that the corrected feature points are stored as an input value in the leaf-shaped outline information DB 153.

Hereinafter, with reference to the accompanying drawings will be described in detail the search method of the plant through the contour of the leaf shape according to the present invention.

First, a system applied to a method for searching plants through the leaf-shaped contour according to the present invention will be described.

1 is a view showing the overall system applied to the method for searching plants through the contour of the leaf-shaped according to the present invention.

As shown in FIG. 1, a system for searching a plant through a leaf-shaped outline according to the present invention includes an information providing server computer 100, a network 200, and a user computer 300.

That is, the information providing server computer 100 collects and studies various plants, and ecological information about the environment of the habitat such as water temperature, hydrogen ion concentration, electrical conductivity and dissolved oxygen amount of the plant's growing place, Morphological information of nutrient and reproductive organs through biological observation, that is, morphological information such as scientific name, lifespan and habit, sketch of leaf shape of each plant, and leaf shape extracted from leaf contours by MPP method It is a terminal that provides a database (DB) of the contour information.

More specifically, the information providing server computer 100 has an interface 110 for exchanging plant-related information in response to a request of the user computer 300 through the network 200, the ecology, form, and Input unit 120 for inputting various types of information such as a leaf-shaped contour by a sketch, and using a MPP (Minimum Perimeter Polygon) technique in the leaf-shaped contour by a sketch inputted through the input unit 120. The extraction unit 125 extracts the feature points of the leaf outline by extracting the feature points of the leaf outline and correcting the rotation, movement, and size of the leaf shape when the feature points of the leaf outline are irregular. An input screen for inputting information on each species related to plants through 120 and an output unit 140 for displaying each species screen, and ecology, shape, and leaves of each aquatic plant input through the input unit 120. The storage unit 140 stores various kinds of information such as sheep contours, and the leaf requested to be searched if there is a search request from the user computer 300 through the contours of a leaf shape of a specific plant, for example, a sketch, a photograph, and a table. Matching unit 130 for searching for a plant having a leaf shape similar to the leaf shape required for the search by comparing the shape and leaf shape existing in the storage unit 140, the interface 110 and each unit ( And a control unit 160 that controls the functions of 120 to 150 to provide the user computer 200 with each type of plant information.

Here, the storage unit 150 is an ecological information DB (151) that has a database of the ecological information on the habitat environment, such as plant names, habitats, growth temperature, hydrogen ion concentration, electrical conductivity and dissolved oxygen amount of each plant and And the morphological information DB (152), which is a database of morphological characteristics of nutritional and reproductive organs, ie scientific names, lifespan, and habits, through plant specimens and biological observations, and sketching leaf shapes of various plants. At the same time, the leaf-shaped contour information DB 153 which extracts the database from the feature points of the leaf-shaped contour by the MPP technique is included.

In addition, the input unit 120 is a means for inputting information related to various plants, although not specifically illustrated in the drawings, a keyboard, a mouse, a digitizer, a scanner, a digital camera, a mobile phone, a PDA (Personal Digital Assistant), and the like. It is a concept including the input means of the same, as is the input unit 320 of the user computer 300 to be described below.

On the other hand, in the present specification, the leaf shape of each plant stored in the leaf-shaped contour information DB 153 is sketched by an illustrator or the like and input through the input unit 120, and the extracting unit 125 with respect to the leaf-shaped sketch. It will be described that the feature points of the leaf-shaped contour extracted through the corresponding.

The network 200 refers to a wireless Internet network such as a TCP / IP protocol that can be wired through a computer and a WAP protocol that can be wirelessly connected.

Meanwhile, the user computer 300 inputs and stores the leaf shape of the plant desired by the user for each plant provided from the information providing server computer 100 through the network 200, and then stores the stored leaf shape. A terminal for uploading and searching, an interface 310 for exchanging information with the information providing server computer 100 through a network 200, and for searching for various plants stored in the information providing server computer 100; In addition to inputting the leaf shape of the plant desired by the user through, for example, sketches, photographs, and the like, the input unit 320 for inputting various kinds of information, and the noise of the photograph and the table input through the input unit 320 A filtering unit 330 for filtering, an output unit 340 for displaying various kinds of information related to plants provided from the information providing server computer 100, and the user computer 300; The storage unit 350, which stores not only the operational information and various software information, but also stores the leaf shape input through the input unit 320, and the various plants stored in the information providing server computer 100. In order to search, the uploading unit 355 uploads the leaf shape stored in the storage unit 350 through the input unit 320, the interface 310 of the user computer 200, and the respective units 320 to 355. The control unit 360 to control the function of the.

At this time, the information providing server computer 100 and the user computer 300 is a computer device that can communicate with the Internet browser that can display web content, and includes a personal computer such as a desktop computer, a notebook computer, etc. At least Windows98 or higher operating system, pentium or higher CPU, 64Mb or more memory is preferably included.

In addition, in the present specification, the user computer 300 is described as the above-described computer device, but as long as it is possible to retrieve various information related to plants provided by the information providing server computer 100, PDA, wireless Internet communication. The mobile phone and IMT-2000 can be used as a matter of course.

Hereinafter, with reference to FIGS. 1 to 9, a search method of a plant through a leaf-shaped outline according to the present invention will be described in detail.

Figure 2 is a flow chart showing a method of searching for plants through the contour of the leaf shape according to the present invention, Figure 3 is a flow chart showing in detail the process of extracting feature points of the leaf shape through the MPP method of step S420 of FIG. 4a and 4b is a view showing the ecological information input form and the input ecological information DB of the plant to be applied to the present invention, Figures 5a and 5b is a form information input form and input form information of the plant to be applied to the present invention It is a figure which shows a specific example of DB.

In addition, Figure 6 is a view showing in detail by applying the flow chart of Figure 3 to the leaf shape, Figure 7 is a view showing the connection between the points and points after merging unnecessary points among the feature points of the extracted leaf-shaped contour. 8 is a diagram illustrating an example of extracting feature points of a leaf contour after correcting rotation, movement, and size when the leaves are not symmetrical, and FIG. 9 is an example of an uploading unit and a display screen of a user computer. It is a figure which shows.

First, the information provider of the information providing server computer 100 collects and studies the plants inhabiting the whole country, as shown in Figure 4a and 5a, the ecological information input form and form information displayed through the output unit 140 In the input form, after inputting the results of each study through the input unit 120, the database is stored in the ecological information DB 151 and the form information DB 152 of the storage unit 150 (S410).

For example, FIGS. 4B and 5B illustrate specific examples of the ecological information DB and the form information DB that are input through the step S410 and are databased.

At the same time, according to the results of the study in step S410, sketching the leaf shape of each species of plant through the input unit 120 and at the same time through the extraction unit 125 through the MPP technique through the feature point of the leaf outline After the extraction, the sketch and the extracted leaf-shaped contour point correspond to each other and are stored in the leaf-shaped contour information DB 153 of the storage unit 150 (S420).

At this time, since the leaves of each plant is a closed-loop in which the start point and the end point of the leaf-shaped contour except the leaf vein are connected to each other, the feature points on the leaf-shaped contour can be extracted.

Therefore, the process of extracting the feature point of the leaf-shaped contour using the MPP technique in step S420 will be described in detail below with reference to FIGS. 3 and 6.

First, in order to extract a feature point from a sketched leaf shape of a specific plant, the leaf shape is divided into cells and converted into a matrix of a simplified form (S421) (see FIG. 6A).

Subsequently, after the closed type internal area is obtained from the transformed matrix, the boundary of the internal area is obtained in the form of four connected points in the internal area (S422).

Next, after obtaining a chain code at the boundary of the inner region, a convex point and a concave point are obtained from the chain code (S423) (see FIG. 6B).

Next, after the first polygon is obtained by connecting the convex points, all of the concave points outside the polygon are removed (S424) (see FIG. 6C).

Subsequently, after removing the remaining convex points and concave points to form a second polygon, the concave points in the convex points in the connected second polygons are removed again (S425).

Thereafter, the step S425 is repeated until there are no more changes, and after removing the points having an angle of 180 degrees, the remaining points are finally used as feature points of the leaf-shaped contour (S426) (see FIG. 6 (d)). .

On the other hand, in the process of extracting the feature point of the leaf-shaped contour by using the MPP method of steps S421 to S426, the convex point (less than 180 degrees) and the concave point (more than 180 degrees) based on the line connecting the points. Since only the feature points of the leaf-shaped outline are found, the extracted feature points of the leaf-shaped outline include many unnecessary points.To simplify these unnecessary points, the angles of each point are calculated to merge the points out of the threshold range. do.

That is, it can be seen that Figure 7 is a diagram connecting points between points after merging unnecessary points among the feature points of the extracted leaf-shaped contour. For example, in Fig. 7, (a) shows before merging the feature points of the extracted leaf-shaped contour, (b) shows after merging unnecessary points, and (c) shows each feature points after merging unnecessary points. It shows the connection.

At this time, the algorithm for determining whether the unnecessary points are merged is shown in Table 1 below.

For example, in order to derive the angles of three points A, B, and C of ΔABC in the algorithm for determining whether unnecessary points are merged in Table 1, the second law of cosines is expressed as in Equation 1 below. Use

(BC의 대응각), b =

(AC의 대응각), c =

(AB의 대응각))Where a =

(Correspondence angle of BC), b =

(Correspondence angle of AC), c =

(Correspondence angle of AB)

의 대응각 a(BC의 대응각)는 아래의 수학식 2와 같다.In this case, if the coordinate of the point B is (x _i , y _i ) and the coordinate of the point C is (x _j , y _j ),

The corresponding angle a (corresponding angle of BC) is given by Equation 2 below.

On the other hand, the leaf shape corresponding to each species of plants to be input should be invariant to the feature point of the contour, that is, the extraction point for movement, size, and rotation. For example, when the leaves are vertical and when they are horizontal, the feature points for the contour of the leaf shape should be the same. In this document, the angle of rotation is corrected based on the two points at the maximum distance from the distance between the dominant points. Also, based on the correction value for each point, the values for movement and size are corrected to the left and right end points and the up and down end points.

That is, in Fig. 8, as shown in (a), the maximum distance of the original leaf shape, i.e., the feature points of the leaf contours represented by the lower point h and the upper point i are irregular, i.e., not symmetrical to one side. If it is biased, as shown in (b), the upper point (i) is corrected to the new upper point (ii) by the rotation angle, and as shown in (c), the new upper point (ii) and the lower point which corrected the rotation angle. The feature point is moved to (h), the size is corrected to the left (j) and the right (k), and the end points of the top, bottom, left and right (h, i, j, k) are corrected as input values.

For example, the algorithm for input value correction (rotation angle correction) of the leaf-shaped outline is shown in Table 2 below.

On the other hand, through the step S420, the sketch and the extracted feature of the leaf-shaped contour correspond to the leaf-shaped contour information DB 153 of the storage unit 150, and then the user of the user computer 300, the information providing server In order to search for a specific plant stored in the leaf-shaped outline information DB 153 of the computer 100, the information server server 100 is connected through the network 200, and the storage unit 350 is input through the input unit 320. The leaves, such as sketches, photos and table books previously stored in the upload through the upload unit 355 (S430).

That is, as shown in FIG. 9, the user of the user computer 300 selects the search 355a of the upload unit 355, finds the leaf shape to be searched through previously stored sketches, pictures, and tables, and then searches. The button 355b is selected to search.

In this case, the display screen 345 of FIG. 9 does not appear at the time of upload, and since the picture and the table except the sketch include noise, the display screen 345 is uploaded after removing the noise through the filtering unit 330.

Here, when uploading the leaves, such as sketches, photographs and tabletops, the extraction of the feature points of the leaf contours to be uploaded is made through the above-described steps S421 to S426, and a detailed description thereof will be omitted here.

Thereafter, the information providing server computer 100 matches the matching unit 130 to search whether or not a plant having a leaf shape similar to the leaf shape uploaded from the user computer 300 exists in the leaf shape contour information DB 153. Calculate the similarity of the leaf shape through (S440).

At this time, Euclidean distance was used to calculate the leaf-like similarity of the step S440, the similarity is calculated as Equation 3 below.

는 사용자 컴퓨터(300)로부터 업로드된 잎모양 윤곽의 특징점의 개수, Di는 u_i에 대한 v_i의 거리로서, 열거의 형태를 가짐)(here,

Is the number of feature points of the leaf-shaped contour uploaded from the user computer 300, Di is the distance of v _i with respect to u _i , having an enumerated form)

That is, the minimum value in the distance between the feature point (u) of the leaf-shaped contour uploaded from the user computer 300 and the feature point (v) of the leaf-shaped contour stored in the leaf-shaped contour information DB 153 of the information providing server computer 100. Calculate using the distance D _i .

Also, the matching unit 130 searches whether the information providing server computer 100 has a leaf shape similar to the leaf shape uploaded from the user computer 300 in the leaf shape contour information DB 153. In order to minimize the search time, the similarity is calculated using a dynamic matching algorithm as shown in Table 3 below.

이고, 잎모양 윤곽정보 DB(153)에 저장된 잎모양의 특징점 개수가

일 때,

<

이면, SMP(v)를 통해 샘플링한 특징점을 추출하여 유사도를 계산할 수도 있다.That is, only one feature point is used on the basis of a straight line (axis) having the longest distance among the feature points of the leaf shape using the feature that the leaf shape of the plant is symmetrical. In addition, if the number of feature points is larger than the maximum value, the similarity is calculated through the sampling function SMP (). For example, the number of feature points of the leaf-shaped contour uploaded from the user computer 300

The number of leaf-shaped feature points stored in the leaf-shaped outline information DB 153 is

when,

<

In this case, the similarity may be calculated by extracting feature points sampled through SMP (v).

Then, in step S440, if a plant having a leaf shape similar to the leaf shape uploaded from the user computer 300 exists in the leaf shape contour information DB 153, the information providing server computer 100 is connected to the network 200. A leaf shape similar to the uploaded leaf shape is displayed on the output unit 340 of the user computer 300 through S450.

That is, as shown in the right side of FIG. 9, a leaf shape similar to a leaf shape uploaded to the display screen 345 displayed through the output unit 340 is displayed.

At this time, if a plant having a leaf shape similar to the leaf shape uploaded from the user computer 300 does not exist in the leaf shape contour information DB 153, the process returns to the step S440.

Lastly, when the user of the user computer 300 displays a similar leaf shape through the display screen 345 of the output unit 340 through step S450, the user selects a corresponding leaf shape, and the user wants to search for the user himself. Various information such as ecology and shape information of a plant having a leaf shape can be confirmed (S460).

For example, as shown in the display screen 345 of FIG. 9, when a leaf shape similar to a leaf shape (eg, 'dry phlegm') uploaded through the uploading unit 355 is displayed on the display screen 345, the user By selecting a leaf shape corresponding to 'dry phlegm', information corresponding to 'dry phlegm' as shown in FIG. 5B can be confirmed.

Although the present invention has been described above according to an embodiment of the present invention, a person skilled in the art to which the present invention belongs has changed and modified within the scope without departing from the technical spirit of the present invention. Of course.

That is, in one embodiment of the present invention is described as searching for a specific plant in the database through the contour of the leaf shape, the specific plant through the contour of the shape of the flower, postcard, leaf, leaf bottom, fruit form, inflorescence and leaf May be searched for, or similarly, insects and animals may be searched.

As described above, according to the search method of the plant through the leaf-shaped contour according to the present invention, after sketching the leaf shape of each plant and extracting the feature point of the leaf-shaped contour by the MPP method and databased, When a user requests a search through the leaf outline of a specific plant, the user can search for and confirm a plant having the leaf shape requested by the user, so that the user does not know the name of the plant to search for and has a specific shape such as a leaf shape. You can easily find and identify the plant without knowing it.

Claims (4)

As a method for searching for various plants through the leaf-shaped contour by using a system including the information providing server computer 100, the network 200 and the user computer 300, (A) The information providing server computer 100 receives the ecological information and the shape information corresponding to each plant through the input unit 120 and stores in the ecological information DB (151) and the shape information DB (152) (S410) After receiving the leaf-shaped sketch of each plant through the input unit 120 and extracting the feature points of the leaf-shaped contour using the MPP method through the extraction unit 125, the sketch and the extracted leaf-shaped contour Mapping the feature points to the leaf-shaped outline information DB 153 (S420) (S410 and S420); (B) a sketch previously stored in the storage unit 350 by the user computer 300 through the input unit 320 to search for a specific plant stored in the leaf-shaped outline information DB 153 of the information providing server computer 100, Uploading a leaf shape such as a photograph and a table through the upload unit 355 (S430); (C) Using the dynamic matching algorithm, the leaf similarity of whether the leaf providing the leaf shape outline information DB 153 similar to the leaf shape uploaded from the user computer 300 is provided by the information providing server computer 100. Searching through the matching unit 130 (S440); (D) If there is a leaf shape similar to the leaf shape uploaded through the step (C), the information providing server computer 100 through the output unit 340 of the user computer 300 through the network 200; Displaying a leaf shape similar to the uploaded leaf shape (S450); And (E) When the leaf shape similar to the uploaded leaf shape is displayed through the output unit 340 through the step (D), the user computer 300 has a leaf shape to be searched by selecting the corresponding leaf shape. Including the information of each species of the plant (S460), In the step (A), When the feature points of the leaf outline are irregular, the leaf shape is corrected by the extracting unit 125 to correct the rotation, movement, and size of the leaf shape, and then the corrected feature points are stored as input values in the leaf shape outline information DB 153. In the step (B), The photograph and the table are after the noise is removed through the filtering unit 330, the search method of the plant through the leaf-shaped contour, characterized in that uploaded through the upload unit (350). The method of claim 1, In the step (A), the process of extracting the feature point of the leaf-shaped contour by using the MPP technique, (A1) cutting the sketched leaves of the specific plant into cells and converting them into a matrix in simplified form (S421); (A2) obtaining a closed inner region from the transformed matrix, and then obtaining a boundary of the inner region in the form of four connected points in the inner region (S422); (A3) after obtaining a chain code at the boundary of the inner region, obtaining convex points and concave points in the chain code (S423); (A4) connecting the convex points to obtain a first polygon, and then removing concave points outside the polygon (S424); (A5) making a second polygon by connecting the convex point and the concave point remaining after the removal, and then removing the concave points in the convex point in the connected second polygon again (S425); And (A6) repeating step (A5) until there is no further change, and after removing the points having an angle of 180 degrees, and finally extracting the remaining points as a feature point of the leaf-shaped contour (S426) Searching method of plant through leaf-shaped contour. delete delete

Images