TWI639138B - Ancient stone image recognition method - Google Patents

Abstract

A method for identifying an image of a historic stone, which allows a user to automatically upload an image of a stone to an image recognition server, and then automatically analyze the RGB value of each pixel in the stone image to obtain the color code of each pixel and pass the color. The similarity degree comparison method analyzes the mineral composition of the stone image, obtains the mineral composition ratio of the stone image by attribution statistics, and further searches for the stone origin information with similar mineral composition ratio through the proportional matching degree comparison method. Thereby, the origin of the stone can be quickly and conveniently determined in a non-destructive manner, so as to facilitate the subsequent stone repairing operation.

Description

Ancient stone image recognition method

The invention relates to an image recognition method, in particular to a method for identifying a monumental stone image.

Since the same stone will have different mineral composition ratios and textures depending on the place of production, when the stone is damaged and needs to be repaired, the primary task is to confirm the origin of the stone and correct it to find the matching stone. To ensure the consistency of the stone after repair, it is based on the current existing stone identification methods, all through the experts to visit the scene to collect the stone samples, and then bring them back to the laboratory for stone analysis, in order to understand the origin of each stone With ingredients.

However, the above method of collecting stone samples for analysis is a destructive analysis, which will destroy the stone, so it is not suitable for environments where monuments, historical buildings, etc. cannot be sampled, and the above methods take a lot of time from collecting to analyzing the results. However, it is impossible for the on-site personnel to immediately know the result. Therefore, after observing the above-mentioned shortcomings, the inventor of the present invention thought that it is necessary to propose a non-destructive and quick and convenient stone identification method, and it is necessary to produce the present invention. .

The object of the present invention is to provide a method for identifying a historic stone image, which can utilize the method of analyzing stone images to identify the composition and origin of the stone in a non-destructive, fast and convenient manner.

In order to achieve the above object, the method for identifying a historic stone image provided by the present invention comprises the steps of: photographing a stone using a photographing device to obtain a stone image, and uploading the stone image to the network through the network. An image recognition server; after receiving the stone image, the image recognition server analyzes the RGB values of each pixel in the stone image to obtain the color code of each pixel; and obtains a plurality of standard color codes of minerals from a database. And color matching the color codes of the pixels to the color similarity of the standard color codes of the minerals. If the similarity is higher than a predetermined value, the pixel is redefined as the corresponding mineral name, and then According to the statistic statistics, the mineral composition ratio of a stone image is obtained; the mineral composition ratio of the plurality of stone producing areas is obtained from the database, and the mineral composition ratio of the stone image is proportional to the mineral composition ratio of the stone producing areas. Yes, to judge the origin of the stone.

According to the image recognition method of the historic stone provided by the invention, the user can automatically analyze the RGB value of each pixel in the stone image through the image recognition server to obtain the color code of each pixel, and Through the color similarity comparison method, the mineral composition of the stone image is analyzed, and the mineral composition ratio of the stone image is obtained by attribution statistics, and further, through the method of proportion matching, the stone origin of the mineral composition is similarly found. Information, in this way, the origin of the stone can be quickly and conveniently determined in a non-destructive manner, so as to facilitate the subsequent stone repair work.

Please refer to FIG. 1 and FIG. 2 , which are schematic diagrams of a connection and a flow chart of a preferred embodiment of the present invention. The method for identifying an image of a historic stone is disclosed. The method for identifying an image of a monumental stone includes the following steps:

Using a photographing device 10 to photograph a stone 20 to obtain a stone image, and simultaneously photographing the color code table image 30 using the photographing device 10 to obtain a color code image, and then the stone image and the color The code list image is uploaded to an image recognition server 50 via the network 40.

After receiving the stone image and the color code table image, the image recognition server 50 analyzes the RGB values of each pixel in the stone image to obtain the color code of each pixel, and obtains a standard color from a database 60. The code table, and the color code table image is color-matched with the standard color code table to obtain a color compensation parameter, and then the color code of the pixels is corrected according to the color compensation parameter.

The standard color code of the plurality of minerals is obtained from the database 60, and the color codes of the pixels are sequentially compared with the standard color codes of the minerals, and if the similarity is higher than a predetermined value, Re-determine the pixel as the corresponding mineral name, and then perform attribution statistics to obtain the mineral composition ratio of a stone image. The color similarity comparison is performed by the following formula: 1-[absolute value (color code of pixel) R ± allowable value - standard color code of mineral R) / 255 + absolute value (color code of pixel G ± allowable value - standard color code of mineral G) / 255 + absolute value (color code of pixel B ± allowable value - mineral The standard color code B) / 255].

Obtaining a mineral composition ratio of the plurality of stone producing areas from the database 60, and comparing the mineral composition ratio of the stone image with the mineral composition ratio of the stone producing areas, thereby determining the origin of the stone, wherein After obtaining the mineral composition ratio of the stone image, a tolerance parameter is first added, and then the database 60 is searched to obtain the mineral composition ratio of the stone producing area, wherein the ratio matching degree is transmitted through the following Formula: Σ{1-[absolute value (mineral composition ratio of stone imagery - mineral composition ratio of stone origin)* mineral composition ratio of stone origin]}.

For a further understanding of the structural features of the present invention, the application of the technical means, and the intended effect, the manner of use of the present invention will be described. It is believed that the present invention may be more deeply and specifically understood as follows:

Please refer to FIG. 1 and FIG. 2 again, and please refer to FIGS. 3 to 5, the user first photographs the stone 20 using the photographing device 10 to obtain the stone image, for example, a granite. The image is uploaded to the image recognition server 50 through the network 40. After receiving the stone image, the image recognition server 50 analyzes the RGB value of each pixel in the stone image to obtain each pixel. The color code, in which, in view of the film image obtained by the filming, is limited by the difference between the shooting device, the field environment and the brightness of the light, and the color of the stone image is easily deviated, so when the film image is obtained, the requirement is The user first prepares the color code table image 30, and simultaneously captures the color code table image 30 by using the image capturing device 10 to obtain the color code table image, and uploads the color code table image to the image recognition server 50, and allows The image recognition server 50 can obtain the standard color code table from the database 60, and compare the color code table image with the standard color code table to obtain the color compensation parameter, and then obtain the color compensation parameter. The color compensation parameter corrects the color code of the pixels, thereby ensuring that the color of the captured stone image is correct, and in order to improve the accuracy of the stone identification, only the stone can be photographed during shooting, and other objects cannot be included. The size of the image is subject to a predetermined format and specification set by a system.

Then, the standard color code of the plurality of minerals corresponding to the mineral component of the stone is obtained from the database, as shown in the example in FIG. 3, and the color code of the pixels obtained by the analysis of the stone image is The color similarity comparison is performed with the standard color code of minerals such as quartz, plagioclase, alkaline feldspar, and biotite. The color similarity comparison is performed by the following formula: 1-[absolute value (color code of pixel R± Allowable value - standard color code of mineral R) / 255 + absolute value (color code of pixel G ± allowable value - standard color code of mineral G) / 255 + absolute value (color code of pixel B ± allowable value - standard of mineral Color code B) / 255], if the similarity is higher than the predetermined value set by the system, the pixel is redefined as the corresponding mineral name, and after all the color codes of all the pixels are compared, the attribution statistics are performed, that is, The proportion of minerals such as quartz, plagioclase, alkaline feldspar, and biotite in the stone image can be obtained.

After obtaining the mineral composition ratio of the stone image, the tolerance parameter as shown in the example in FIG. 4 is first added, and then the search is performed in the database 60, thereby obtaining a mineral composition ratio with the stone image. The mineral composition ratio of the plurality of stone producing areas is close to, and the mineral composition ratio of the stone image is compared with the mineral composition ratio of the stone producing areas, and the ratio matching degree is performed by the following formula: Σ{ 1-[absolute value (mineral composition ratio of stone image - mineral composition ratio of stone origin) * mineral composition ratio of stone origin]}, by which the results can be compared according to the matching degree, and the example shown in the fifth example is obtained. The list of stone origins is returned to the user, so that the user can immediately and clearly judge the origin of the stone, so as to facilitate the subsequent stone repair work.

Hereafter, the features of the present invention and its achievable efficacy are stated as follows:

According to the image recognition method of the historic stone of the present invention, the user can automatically analyze the RGB value of each pixel in the stone image through the image recognition server 50 by taking the image of the stone image, and obtain the color code of each pixel. The color similarity comparison method analyzes the mineral composition of the stone image, obtains the mineral composition ratio of the stone image by attribution statistics, and further searches for the stone origin information with similar mineral composition ratio through the proportional matching method. Thereby, the origin of the stone can be quickly and conveniently determined in a non-destructive manner, so as to facilitate the subsequent stone repairing operation.

In summary, the present invention has excellent advancement and practicability in similar products, and at the same time, the technical materials of such structures are frequently investigated at home and abroad, and the same structure is not found in the literature. The invention already has the invention patent requirements, and the application is filed according to law.

However, the above-mentioned ones are merely preferred embodiments of the present invention, and the equivalent structural changes of the present invention and the scope of the claims are intended to be included in the scope of the present invention.

10‧‧‧Photographing device

20‧‧‧ Stone

30‧‧‧ color code table image

40‧‧‧Network

50‧‧‧Image recognition server

60‧‧‧Database

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a schematic view of the linkage of a preferred embodiment of the present invention. Figure 2 is a schematic flow diagram of a preferred embodiment of the present invention. Figure 3 is a diagram of a standard color code for minerals in accordance with a preferred embodiment of the present invention. Figure 4 is a table of allowable value parameters for a preferred embodiment of the present invention. Figure 5 is a list of stone origins of a preferred embodiment of the present invention.

Claims (5)

A method for identifying an image of a historic stone material, comprising the steps of: photographing a stone with a photographing device to obtain a stone image, and uploading the stone image to an image recognition server through a network; the image recognition servo After receiving the stone image, the RGB value of each pixel in the stone image is analyzed to obtain the color code of each pixel; the standard color code of the plurality of minerals is obtained from a database, and the color code of the pixels is obtained. Color similarity comparison with the standard color code of the minerals is sequentially performed. If the similarity is higher than a predetermined value, the pixel is redefined as the corresponding mineral name, and then the attribution statistics are performed to obtain a mineral of a stone image. Composition ratio; The mineral composition ratio of the plurality of stone producing areas is obtained from the database, and the mineral composition ratio of the stone image is compared with the mineral composition ratio of the stone producing areas to determine the origin of the stone. According to the method for identifying an image of a historic stone according to claim 1, wherein, when the stone is photographed, a color code image is simultaneously captured to obtain a color code image, and the image recognition server is A standard color code table is obtained in the database, and the color code table image is color-matched with the standard color code table to obtain a color compensation parameter, and the color code of the pixels is corrected according to the color compensation parameter. According to the method for identifying the image of the historic stone according to the first aspect of the patent application, after obtaining the mineral composition ratio of the stone image, a tolerance parameter is added first, and then the database is searched to obtain the stone. The proportion of mineral composition in the place of origin. According to the method for identifying an image of a historic stone according to claim 1, wherein the color similarity comparison is performed by the following formula: 1-[absolute value (color code of pixel R±allowed value-standard color code of mineral) R) / 255 + absolute value (pixel color code G ± allowable value - mineral standard color code G) / 255 + absolute value (pixel color code B ± allowable value - mineral standard color code B) / 255]. According to the image recognition method of the historic stone according to Item 1 of the patent application scope, the ratio matching degree is performed by the following formula: Σ{1-[absolute value (mineral composition ratio of stone image-mineral composition of stone origin) Proportion) * Mineral composition ratio of stone origin]}.