KR20090011423A - Method of discern home production agricultural products with portable photograph device - Google Patents

Abstract

A method for distinguishing a domestic agricultural product using a portable multimedia apparatus is provided to distinguish a domestic agricultural product and obtain a result in a same place by using a portable multimedia apparatus. A method for distinguishing a domestic agricultural product using a portable multimedia apparatus comprises the following steps. Multimedia information of a domestic agricultural product is collected and analyzed. A specific database of multimedia information about the domestic agricultural product is stored in a server. Multimedia information of comparison group is collected using a portable multimedia apparatus such as a camcorder, a camera, and a mobile phone with a camera.

Description

Method of discern home production agricultural products with portable photograph device}

The present invention comprises the steps of: 1. collecting and analyzing the image information of the domestic agricultural products and constructing a unique database of domestic agricultural image information on the server; 2. collecting image information of the control agricultural product using a portable imaging device and transmitting the image information to the server through a disk or wired or wireless; 3. receiving image information of the control agricultural product from the server and analyzing the control data with the control data to be compared with the unique database of 1; 4. In the server, the judging step of comparing the control data of 3 with the unique database information of 1 to determine whether to produce domestically produced agricultural products; It relates to a method for discriminating domestic agricultural products using a portable imaging device, characterized in that made.

Currently, Korea is importing a lot of foreign agricultural products in the age of import liberalization.

However, since these imported agricultural products are generally of lower quality than the domestic ones, but the price is cheaper, many of them are intentionally deceived when they are sold. The government mandates the labeling of domestic agricultural products by the relevant laws, Because of hiding or selling a mixture of domestic and foreign agricultural products are serious damage to the country of origin.

Currently, research on accurate and rapid quality evaluation methods for these imported agricultural products is being conducted. In addition to the traditional physicochemical method, the technology that can quickly determine their quality or authenticity in the field is a non-destructive technology, near infrared spectroscopy, Electronic nose, X-ray fluorescence, etc. are used a lot.

Non-destructive technology using the near-infrared spectroscopy has been widely used because it can easily measure the sugar content of fruit, the yield of cereals, the content of amylose, the fat, protein and solids content of milk and dairy products. Near-infrared spectroscopy is a method of irradiating near-infrared light in the 700 ~ 2500nm region with a near-infrared spectrometer and analyzing the reflected and transmitted spectra. It is characterized by the absorption band of overtone and combination of fundamental frequency and its strength is relatively weak. Until chemometrics were developed, they were difficult to distinguish from noise and were not used in actual analysis, but were first used by Karl Norris to analyze the moisture, fat and protein of agricultural products in agriculture. Due to the development of the application is expanding its application to many fields.

In addition to the spectroscopic method, various determination methods, such as DNA analysis, have been studied. These methods are expensive and difficult to install and use, requiring specialized knowledge, which makes it very difficult to be applied by merchants in the actual market. There is this.

In order to solve the above problems, an object of the present invention is to provide a method for discriminating domestic agricultural products using a portable imaging apparatus that analyzes the appearance of agricultural products using a portable imaging apparatus to determine imported and domestic agricultural products.

In particular, it is an object of the present invention to provide a method for discriminating domestic agricultural products using a portable imaging apparatus, in which a request for determination and a reply to a result can be realized immediately in the field using the imaging apparatus of a mobile phone.

In order to achieve the above object, the present invention comprises the steps of: 1. collecting and analyzing the image information of the domestic agricultural products and constructing a unique database of the domestic agricultural image information on the server; 2. collecting image information of the control agricultural product using a portable imaging device and transmitting the image information to the server through a disk or wired or wireless; 3. receiving image information of the control agricultural product from the server and analyzing the control data with the control data to be compared with the unique database of 1; 4. In the server, the judging step of comparing the control data of 3 with the unique database information of 1 to determine whether to produce domestically produced agricultural products; A technical gist of the method for discriminating domestic agricultural products using a portable imaging apparatus, characterized in that the present invention is made.

Wherein the determination step 4 is characterized in that the fuzzy inference of a specific value in the unique database of the first step, and then compares the fuzzy inferred result with the control data of the third step to determine whether or not domestic production of agricultural products It is preferable to use a method for discriminating domestic agricultural products using an imaging device.

The specific value data analyzed and fuzzy inferred in steps 1 and 3 may be a method for discriminating domestic produce using a portable imaging apparatus, wherein the specific value data comprises morphological information of the produce.

In addition, the specific value data is preferably a method for discriminating domestic agricultural products using a portable imaging device, characterized in that further comprising the contrast information of agricultural products.

In addition, it is preferable that the specific value data is a method for discriminating domestic agricultural products using a portable imaging apparatus, characterized in that the color information of agricultural products is further included.

In addition, the morphometric features of the produce are features of a minimum rectangle, and include the longest extension information in the horizontal direction and the longest extension information in the vertical direction with respect to the principal axis of the produce. Obtaining ratio of agricultural product area information and rectangular area (area area of object / area of minimum rectangle area); Characteristic of the minimum bounding rectangle, the length information of the long axis in the ellipse in the image, the length information of the short axis in the ellipse and the area information of the ellipse; As geometric features, the perimeter information of the image, the ratio of the perimeter of the produce to the perimeter (breath_to_peri) and the information of the circularity; It is preferable that the method is a method for discriminating domestic produce using a portable imaging apparatus, characterized in that the ratio information of height and width is obtained.

In addition, the contrast information of the agricultural product is a gray level (gray level) -related characteristics, gray_mean information of brightness, gray_std information of brightness, dispersion information of brightness and CV CV (coefficient of variation) information Acquiring; As a histogram-related feature, it is desirable to be a method for discriminating domestic produce using a portable imaging device, characterized in that the smallest contrast value is obtained from the histogram and the largest contrast value is obtained from the histogram.

In addition, the color information of the produce is the average of the red color (red_mean), the standard deviation of the red (red_std), the center value (red_center) through the red k-means, the average through the red k-means (red_k_mean), Green mean (green_mean), green standard deviation (green_std), green mean k-means center (green_center), green mean k-means (green_k_mean), blue mean (blue_mean), Domestic produce using a portable imaging device, characterized by obtaining information on the standard deviation of blue color (blue_std), the center value through blue k-means (blue_center), and the average through blue k-means (blue_k_mean) It is desirable to be a discrimination method.

According to the above invention, there is an advantage that a method for discriminating domestic produce using a portable imaging apparatus for determining the imported and domestic produce by analyzing the appearance of the produce using the portable imaging apparatus.

In particular, there is an advantage that a method for discriminating domestic agricultural products using a portable imaging apparatus, in which a request for determination and a reply of a result can be immediately realized in the field using the imaging apparatus of a mobile phone.

Hereinafter, the present invention will be described with reference to the accompanying drawings. In describing the present invention, when it is determined that the detailed description of the related well-known technology or configuration may unnecessarily obscure the gist of the present invention, the detailed description thereof will be omitted. will be.

The terms to be described below are terms defined in consideration of functions in the present invention, and may be changed according to intentions or customs of users or operators, and the definitions should be made based on the contents throughout the specification for describing the present invention.

1 is a block diagram of a method for discriminating domestic produce using a portable imaging apparatus according to the present invention.

The present invention relates to a method of determining whether a domestic product of a farm product is returned by transmitting a video information of the farm product using a camcorder, a camera, or a mobile phone in which the camera is embedded.

Referring to Figure 1, the present invention is largely 1. the step of building a unique database of domestic agricultural product image information on the server, and 2, transmitting the image information of the control agricultural product using a portable image device and 3. the control agricultural products 4. It consists of analyzing the video information of the judgment step.

Step 1 is a step of collecting the image information of the domestic agricultural products and analyzing them, and constructing a unique database of the domestic agricultural image information on the server.

This step is to collect and process the image information of domestic agricultural products, collect and process the information so that a specific value can be inferred and produce a database and build it on a server to store and process it.

At this time, the data information of the image information is morphological information, contrast information, color information of agricultural products.

Morphological information of the agricultural product is collected by dividing the features of the minimum rectangle (minimum rectangle), the minimum bounding rectangle (minimum bounding rectangle), geometric features (geometric features),

The minimum square region is an area that can be effectively used when extracting features of relatively round agricultural products, such as amber, where 'minimum' is a minimum region that includes agricultural objects present in the image. It means the rectangular area where only the object is extracted without including the area.

Horizontal longest extension information (horizontal longest extension information) with respect to the principal axis of the produce means the X-axis, that is, the horizontal length, of the agricultural object obtained using the X and Y coordinates of the minimum rectangular region. .) And the longest extension information in the vertical direction (the longest extension information in the vertical direction means the Y axis, or length, of the agricultural object obtained by using the X and Y coordinates of the smallest rectangular area) and the agricultural product area information ( The area information of agricultural products refers to the size of pure agricultural object in the minimum square area) and the ratio of the rectangular area (the area of the object / the area of the minimum square area: the ratio of the rectangular area is the area of the agricultural object and the area of the minimum square area). Divided, it represents the percentage of the area of pure agricultural objects described above contained in the minimum square area, which is the product of the longest extension in the horizontal and vertical directions. It will be obtained.

The minimum bounding rectangle is a region that can effectively extract feature values when extracting feature information of an oval agricultural product, such as carrots. After the preprocessing process, the agricultural object is extracted by using the contour tracking algorithm, which is an image processing algorithm that remembers the circumference coordinates in the image until it follows the circumference of the object. The length information of the long axis of, the length of the short axis of the ellipse and the area information of the ellipse are obtained.

The geometric features refer to the perimeter information of the image obtained from the shape, the perimeter of the produce, and the perimeter of the square (the square perimeter is the perimeter of the object obtained from the square image). Ratio (breath_to_peri) information, information on the degree of roundness, and information on the ratio of height and width (proportion of height and width is simply the agricultural product in the smallest rectangular area of the object). The ratio of height to width is defined as 'the vertical length of the minimum square area / the horizontal length of the minimum square area', for example, for a relatively round amber, the rectangular area is almost square. By the shape of, the value will be close to 1.0. On the other hand, in the case of carrot, since it is close to the ellipse, the value is equal to 1.0 by having the rectangular rectangular area. It's hard.)

Contrast information of the agricultural product is gray level (gray level) -related characteristics, gray_mean information of brightness, standard deviation (gray_std) information of brightness and dispersion information of brightness and

CV (coefficient of variation) (CV is one of the characteristics of the contrast information of agricultural products, which is obtained by dividing the standard deviation and average value of agricultural contrast in the gray level area. Contrast is also defined as mean.) Obtain information; As a histogram-related feature, the smallest contrast value in the histogram and the largest contrast value in the histogram are obtained.

The color information of the agricultural product is obtained by separating the color of the image into RGB to obtain the mean, standard deviation, the center value through k-means and the mean value through k-means.

Thus, from the color information of the produce, the red mean (red_mean), the red standard deviation (red_std), the red mean k-means (red_center), and the red mean k-means (red_k_mean) and; The average of green color (green_mean), the standard deviation of green color (green_std), the center value through green k-means (green_center), and the average through green k-means (green_k_mean); Information about the mean of blue color (blue_mean), the standard deviation of blue color (blue_std), the center value through blue k-means (blue_center), and the mean through blue k-means (blue_k_mean) are obtained.

In the present invention, the data of the fuzzy inference is prepared using the above information and built in the server.

The second step of the present invention is a step of collecting image information of the control agricultural products using a portable imaging device such as a camcorder, a camera or a mobile phone with a built-in camera, and transmits them to the server through a disk or wired or wireless.

The control agricultural product is a control to be determined whether or not domestic, according to the present invention has the advantage that the inspection is performed immediately in the field because it is only necessary to shoot and transmit the image without having to carry the agricultural product at the inspection site.

In particular, when using a mobile phone with a built-in camera can be immediately sent to the server in the field pictures by wireless communication, the result is an advantage that can be immediately received by text or voice.

In the third step of the present invention, the server receives the image information of the control agricultural product and analyzes the control data to be compared with the unique database of the first.

The analysis of the control data in this step is the same as the analysis item in the first step, of course.

The fourth step of the present invention is a determination step in which the server compares the control data of the 3 with the unique database information of the 1 to determine whether the agricultural product is domestic.

In the determining step, the fuzzy inference value of the unique database of the first step is compared with the control data of the third step to determine whether the agricultural product is domestic.

Fuzzy inference is a technique of inferring what the current object is by applying the given inference rule using the degree of belonging of the fuzzy membership function using the feature information of each object. The membership function largely designed the membership function for color information of agricultural products and the head thickness ratio membership function for root thickness of agricultural products.

The fuzzy affiliation section for color information is divided into three sections and consists of Dark, Noraml, and Bright. In the dark section, the belonging product of dark color produce with low color information value will appear high, and in the normal section, the brighter produce will show high belonging rate. In addition, in the Bright section, abundance of agricultural products with much bright colors will appear high. When the belonging degree is calculated for each section, the agricultural products belong to the section with the highest belonging degree.

The head function ratio of the root thickness to the root thickness of agricultural products was designed into three affiliation sections, divided into Small, Similar, and Big. Small section shows high membership when head and root thickness is small, and similar section shows high membership when head and root thickness are similar. Big section shows high membership when head and root thickness is big. When the belonging section of the agricultural product is determined, the agricultural product is inferred using the inference rule. Inference rules were established based on the identification information presented by the National Agricultural Products Quality Management Service, and agricultural products were classified.

Hereinafter will be described with an embodiment of the present invention.

Example 1 Comparison of Characteristics of Domestic Pumpkins and Imported Pumpkins

For comparison of features, 50% Gaussian noise was applied to images of domestic pumpkin, imported pumpkin, domestic carrot, imported carrot 1, and imported carrot 2.

1. The results of morphological information comparison are shown in the following table.

Domestic Pumpkin Imported pumpkin Round 715 709 The ratio between the perimeter of the object and the square 0.7788 0.7860 Roundness 0.9997 0.9989 Ratio of height to width 1.0131 0.9867 The main direction of the object 3 3 Longest extension in the horizontal direction 228 227 Longest extension in the vertical direction 231 224 Ratio of area of object to area of rectangular area 0.7726 0.7862 The length of the long axis in the oval 120 119 The length of the short axis in the oval 108 107 The area of the object 40694 39981 Distance between long axis and short axis 105035 103039

Domestic carrot Imported Carrot 1 Imported Carrot 2 Round 621 584 530 Ratio between circumference and square circumference 0.7685 0.7392 0.6279 Roundness 0.6816 0.6503 0.5219 Height to width ratio 0.4175 0.3435 0.7016 The main direction of the object 0 0 One Longest extension in the horizontal direction 285 294 248 Longest extension in the vertical direction 119 101 174 Ratio of area to area of rectangular area 0.6170 0.5946 0.2705 The length of the long axis in the oval 155 148 143 The length of the short axis in the oval 43 38 26 The area of the object 20928 17659 11674 Distance between long axis and short axis 81895 79963 73610

2. Comparison of feature values of domestic carrots and imported carrots

Domestic Pumpkin Imported pumpkin Gray Level  Average 57 94 Gray Level  Standard Deviation 0.8507 0.34188 Gray Level  Dispersion 0.7237 0.11688 Gray Level CV 0.0149 0.0036 Red Level  Average 48 95 Red Level 0.0004 -0.0010 Green Level 62 99 Green Level 0.0014 0.0010 Blue Level 63 89 Blue Level 0.0019 0.0015

Domestic carrot Imported Carrot 1 Imported Carrot 2 Gray Level  Average 139 135 142 Gray Level  Standard Deviation 0.5634 0.7332 0.2082 Gray Level  Dispersion 0.3174 0.5376 0.04338 Gray Level CV 0.0040 0.0054 0.0014 Red Level  Average 186 207 231 Red Level -0.0025 -0.0003 -0.0044 Green Level 134 119 116 Green Level -0.0007 0.0002 0.0014 Blue Level 98 81 79 Blue Level 0.0008 0.0006 0.0041

3. Results

The experimental environment was implemented with Borland C ++ Builder 6.0 on Pentium-IV PC, and Fuzzy Toolbox of MATLAB R2006b was used for the fuzzy membership function. A total of 250 samples were tested by converting white noise, brightness, contrast, and slope / size into 320 * 240 Korean and imported pumpkin and carrot images, respectively.

3.1 Detection of Agricultural Products in Images Using Proposed Agricultural Detection Algorithm

The results of applying the proposed agricultural product detection algorithm to a total of 50 images by converting white noise, brightness, contrast, and tilt / size to domestic pumpkins and imported pumpkins are as follows.

Table 1. Domestic and imported pumpkin detection results

Detect / Total Detection rate Domestic pumpkin 50/50 100.0% Imported pumpkin 50/50 100.0%

The results of applying the proposed agricultural product detection algorithm to the domestic carrots, imported carrots 1, and imported carrots 2 by converting white noise, brightness, contrast, and tilt / size into a total of 50 images, respectively, are as follows.

Table 2. Domestic and Imported Pumpkin Detection Results

Detect / Total Detection rate Domestic carrot 50/50 100.0% income Carrot1 50/50 100.0% income Carrot2 50/50 100.0%

3.2 Classification of Agricultural Products Using Characteristics and Fuzzy Membership

The results of applying the proposed agricultural product classification algorithm to the images of domestic pumpkins and imported crops detected through the proposed agricultural product detection algorithm are as follows.

Table 3. Domestic and Imported Pumpkin Classification Results

Direct classification Proposed Classification Method Classification rate Domestic pumpkin 50/50 50/50 100.0% Imported pumpkin 50/50 50/50 100.0%

The results of applying the proposed agricultural product classification algorithm to the images of domestic carrots, imported carrots 1, and imported carrots 2 detected through the proposed agricultural product detection algorithm are as follows.

Table 4. Domestic and Imported Carrot Detection Results

Direct classification Proposed Classification Method Classification rate Domestic carrot 50/50 50/50 100.0% income Carrot1 50/50 50/50 100.0% income Carrot2 50/50 50/50 100.0%

According to the present invention, it is possible to determine domestic crops and imported crops by morphological analysis as in the above-described embodiment, and when the results are transmitted using a portable imaging apparatus, it is advantageous to easily determine whether the agricultural products are imported on the spot. It can be seen.

Embodiments shown for the purpose of the present invention described above are only one embodiment in which the present invention is embodied, and as shown in the drawings, it can be seen that various forms of combinations are possible to realize the gist of the present invention.

Therefore, the present invention is not limited to the above-described embodiments, and various changes can be made by any person having ordinary skill in the art without departing from the gist of the present invention as claimed in the following claims. It will be said that the technical spirit of this invention is to the extent possible.

1 is a block diagram of a method for discriminating domestic produce using a portable imaging device according to the present invention

Claims (8)

1. collecting and analyzing the image information of the domestic agricultural products and constructing a unique database of the domestic agricultural image information on the server; 2. collecting image information of the control agricultural product using a portable imaging device and transmitting the image information to the server through a disk or wired or wireless; 3. receiving image information of the control agricultural product from the server and analyzing the control data with the control data to be compared with the unique database of 1; 4. In the server, the judging step of comparing the control data of 3 with the unique database information of 1 to determine whether to produce domestically produced agricultural products; Domestic agricultural product identification method using a portable imaging device, characterized in that made. The method of claim 1, wherein the determining step of 4 Determining whether a domestic product is domestic by using a portable imaging apparatus after fuzzy inference of a specific value in the unique database of the first step, and then comparing the fuzzy inferred result with the control data of the third step. Way. The method of claim 2, wherein the specific value data analyzed and fuzzy inferred in steps 1 and 3 A method for discriminating domestic produce using a portable imaging device, characterized in that the morphological information of the produce. The method of claim 3, wherein the specific value data A method for discriminating domestic produce using a portable imaging device, characterized by further comprising contrast information of the produce. The method of claim 4, wherein the specific value data A method for discriminating domestic produce using a portable imaging device, characterized in that the color information of the produce is further included. The method of claim 3 wherein the morphometric features of the produce As a characteristic of the minimum rectangle, For the principal axis of the produce The longest extension in the horizontal direction The longest extension in the vertical direction Agricultural product area information Obtaining ratio (width / height) information of the blind area; As a feature of the minimum bounding rectangle, The length of the long axis in the ellipse Information about the length of the short axis Obtain area information of an ellipse; As geometric features, Perimeter information of the image Information about the ratio between the perimeter of crops and the squares (breath_to_peri) Information of roundness; A method for discriminating domestic produce using a portable imaging device, characterized in that the ratio information of height and width is obtained. The contrast information of the agricultural product is As a gray level related feature, The average brightness (gray_mean) information Standard deviation of brightness (gray_std) Brightness information and Obtain CV CV of coefficient of variation information; As a histogram related feature, The smallest contrast value in the histogram A method for discriminating domestic produce using a portable imaging device, characterized by obtaining the largest contrast value in a histogram. The color information of the produce of claim 5 Red color mean (red_mean) Red standard deviation (red_std) Center value through red k-means (red_center) Mean through red k-means (red_k_mean) Average of green color (green_mean) Green standard deviation (green_std) Center value through green k-means (green_center) Average through green k-means (green_k_mean) Mean of blue color (blue_mean) Blue standard deviation (blue_std) Center value through k-means of blue color (blue_center) Mean through blue k-means (blue_k_mean) Domestic agricultural product identification method using a portable imaging device, characterized in that to obtain information about.

Images