A Chinese painting recognition method based on ink analysis

Abstract

The present invention discloses a Chinese painting recognition method based on ink analysis. First, a basic database is established. The data of Chinese paintings from modern times and before in China is collected using the Internet, and for some Chinese painting works that do not exist online, sample images and sample data are obtained through a scanning device. This Chinese painting recognition method based on ink analysis associates the era with the ink spectrum data and multi-light source detection data to quickly locate the historical years of the Chinese painting. By analyzing and comparing the character with the types of ink styles of the figures in the Chinese painting, the painting style of the Chinese painting is accurately locked. Finally, the two are historically associated to achieve the purpose of quickly judging the Chinese painting. By preferentially establishing a basic database and a real-time database, the data required for recognition is collected, which plays a key role in quickly recognizing Chinese paintings. A log is established in the basic database to save the classification range, improving the learning ability of recognition and further ensuring the accuracy of recognition.

Description

Technical Field

[0001] The present invention relates to the technical field of Chinese painting recognition, and specifically provides a Chinese painting recognition method based on ink analysis. Background Art

[0002] Chinese painting works and calligraphy works are important components of Chinese culture, with unique art forms, forming a self - contained and distinctive system in the world of fine arts. In recent years, with the rapid development of technologies such as computers and multimedia, the digital images of Chinese painting works and calligraphy works have been increasing day by day. Therefore, how to effectively retrieve Chinese painting images and calligraphy images has attracted more and more attention. Chinese calligraphy and painting artworks condense the essence of thousands of years of Chinese cultural history and are also the soul of Chinese culture and art. However, so far, due to serious defects in the identification technology, the market transactions of artworks such as calligraphy and paintings are chaotic, and the unhealthy trends in the industry cannot be prohibited. The true artistic value of artworks such as calligraphy and paintings is difficult to play, and the entire art market is difficult to plan as a whole. Although there are individual technologies involved, they are difficult to have a practical impact on the transactions in the Chinese art market fundamentally. Using "identity cards for calligraphy and paintings" and "filing certificates" to solve the problem of difficult market appraisal, because they do not deviate from the market positioning scope of traditional appraisal, the appraisal is still defined as the responsibility of the purchaser. These certificates and online inquiries are just decorative elements in the market for specific objects of calligraphy and painting artworks. Moreover, these technical operations are under the control of human factors, and the "certificates" can only constitute the instructions for calligraphy and painting objects. In practice, the factual basis for identification is unclear and the evidence is insufficient. The calligraphy and painting identification device technology is an advanced market - oriented technology in the identification application of the Chinese calligraphy and painting art market. It first realizes the seamless connection between the pre - filed information of calligraphy and paintings and the physical objects of calligraphy and paintings, and is a cutting - edge technology for autonomous identification and precise delivery of market transaction information, and has achieved good results in market applications. The existing calligraphy and painting identification technologies cannot directly and quickly identify Chinese paintings through ink analysis. Therefore, a Chinese painting recognition method based on ink analysis is proposed. Summary of the Invention

[0003] (1) Technical Problems to be Solved

[0004] Aiming at the deficiencies of the existing technology, the present invention provides a Chinese painting recognition method based on ink analysis, which can directly and quickly identify Chinese paintings through ink analysis, etc., and solves the problem that Chinese paintings cannot be identified through ink analysis.

[0005] (2) Technical Solutions

[0006] To achieve the above - mentioned purpose of directly and quickly identifying Chinese paintings through ink analysis, the present invention provides the following technical solutions: A Chinese painting recognition method based on ink analysis, including the following steps:

[0007] 1) First, establish a basic database. Use the Internet to collect data on Chinese traditional paintings from modern times and before. For those traditional Chinese paintings without data on the Internet, scan them with scanning equipment to obtain sample images and sample data.

[0008] 2) Then, establish a real-time database. Place the traditional Chinese painting to be recognized in the scanning equipment for image acquisition and data acquisition, and store the obtained acquisition data inside the real-time database.

[0009] 3) Verify the data collected in step 1 with professionals, and set classification intervals for the era and figures inside the basic database. Finally, classify and store the verified data.

[0010] 4) Establish an overall image grid board for recognizing traditional Chinese paintings. According to the screen for observing the image, establish and display on the display screen system a grid template that evenly divides the screen into several equal parts.

[0011] 5) Transmit the images collected inside the real-time database to the display screen. Through the image recognition and analysis module, compare the ink style of the traditional Chinese painting to be recognized with the ink style data in the figure classification interval collected in the basic database, and then output the type of the figure ink style of the traditional Chinese painting to be recognized.

[0012] 6) Make a data judgment on the ink data of the traditional Chinese painting to be recognized collected inside the real-time database and the existing ink data of traditional Chinese paintings collected in the basic database, and output the type of the ink of the traditional Chinese painting to be recognized.

[0013] 7) Overlap the judgment information obtained in steps 5 and 6 in terms of historical relationship to make the final judgment on the recognition of the traditional Chinese painting.

[0014] Preferably, the sample data and the data of the traditional Chinese painting to be recognized in steps 1 and 2 include ink spectrum data and multi-source detection data.

[0015] Preferably, after the ink spectrum data and the multi-source detection data are processed, one or more characteristic data are obtained.

[0016] Preferably, the characteristic data includes data formed by spectral data, peak values associated with the spectral data, fluorescence quantization data, infrared light quantization data, ultraviolet light quantization data, extended light quantization data, color, type, and creation time.

[0017] Preferably, in step 3, inclusion processing is performed between the eras and the classification intervals of the characters. Each character data is associated within the era interval. By setting the association between the era and the ink spectrum data and the multi-light source detection data, the historical years of the Chinese painting to be recognized are quickly located. By analyzing and comparing the character with the types of ink styles of the characters in the Chinese painting, the painting style of the Chinese painting to be recognized is accurately locked. Finally, the two are historically associated to achieve the purpose of quickly judging the Chinese painting.

[0018] Preferably, in step 4, when the grid template is used to click on any grid square of the template with a finger each time, the effect of the grid is displayed. The corresponding grid boxes are not limited in shape, and the markings are not limited in color.

[0019] Preferably, in step 7, a log storage classification interval is established in the basic database.

[0020] Preferably, in step 7, the data of the Chinese painting to be recognized and the recognition steps are recorded to generate a log and saved in the log storage classification interval of the basic database. By preferentially establishing the basic database and the real-time database, the data required for recognition is collected, which plays a key role in quickly recognizing the Chinese painting. And a log storage classification interval is established in the basic database to save the recognition historical data, improving the learning nature of recognition and further ensuring the recognition accuracy.

[0021] (III) Beneficial Effects

[0022] Compared with the prior art, the present invention provides a method for recognizing Chinese paintings based on ink analysis, having the following beneficial effects:

[0023] 1. For the method for recognizing Chinese paintings based on ink analysis, by setting the association between the era and the ink spectrum data and the multi-light source detection data, the historical years of the Chinese painting to be recognized are quickly located. By analyzing and comparing the character with the types of ink styles of the characters in the Chinese painting, the painting style of the Chinese painting to be recognized is accurately locked. Finally, the two are historically associated to achieve the purpose of quickly judging the Chinese painting.

[0024] 2. For the method for recognizing Chinese paintings based on ink analysis, by preferentially establishing the basic database and the real-time database, the data required for recognition is collected, which plays a key role in quickly recognizing the Chinese painting. And a log storage classification interval is established in the basic database to save the recognition historical data, improving the learning nature of recognition and further ensuring the recognition accuracy. Specific Embodiments

[0025] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0026] A Chinese painting recognition method based on ink analysis, comprising the following steps:

[0027] 1) First, establish a basic database, collect data of Chinese paintings from modern times and before in China using the Internet, and scan Chinese paintings without existing data on the Internet through a scanning device to obtain sample images and sample data;

[0028] 2) Then, establish a real-time database, place the Chinese painting to be recognized in the scanning device for image acquisition and data acquisition, and store the obtained acquisition data inside the real-time database. The sample data and the data of the Chinese painting to be recognized in steps 1 and 2 include ink spectrum data and multi-light source detection data. After the ink spectrum data and multi-light source detection data are processed, one or more characteristic data are obtained. The characteristic data include spectrogram data, peak values associated with the spectrogram data, fluorescence quantization data, infrared light quantization data, ultraviolet light quantization data, extended light quantization data, data formed by color, type, and creation time;

[0029] 3) Verify the data collected in step 1 with professionals, set time-based and artist-based classification intervals inside the basic database, and finally classify and store the verified data. Inclusion processing is performed between the time-based and artist-based classification intervals, and each artist-based data is associated within the time-based interval;

[0030] 4) Establish an overall image grid board for recognizing Chinese paintings. According to the screen for observing the image, establish and display a grid template that evenly divides the screen into several equal parts in the display screen system. When any grid square of the template is tapped with a finger each time, the effect of the grid is displayed. The corresponding grid boxes are not limited in shape, and the markings are not limited in color;

[0031] 5) Transmit the images collected inside the real-time database to the display screen. After comparing the ink style of the Chinese painting to be recognized with the ink style data in the artist-based classification interval collected in the basic database through the image recognition and analysis module, output the types of the ink styles of the Chinese painting to be recognized;

[0032] 6) Make a data judgment on the ink data of the Chinese painting to be recognized collected inside the real-time database and the existing ink data of Chinese paintings collected in the basic database, and output the types of the ink of the Chinese painting to be recognized;

[0033] 7) Overlap the obtained judgment information in steps 5 and 6 for historical relationships, make a final Chinese painting recognition judgment, establish a log storage classification interval in the basic database, record the data of the Chinese painting to be recognized and the recognition steps to generate a log and retain it in the log storage classification interval in the basic database.

[0034] The beneficial effects of the present invention are as follows: For the Chinese painting recognition method based on ink analysis, by associating the epochality with the ink spectrum data and multi-light source detection data, the historical years of the Chinese painting to be recognized can be quickly located. By analyzing and comparing the character with the types of ink styles of the figures in the Chinese painting, the painting style of the Chinese painting to be recognized can be accurately locked. Finally, the two are historically associated to achieve the purpose of quickly judging the Chinese painting. By preferentially establishing a basic database and a real-time database, the data required for recognition are collected, which plays a key role in quickly recognizing the Chinese painting. And a log is established in the basic database to save the classification interval, and the historical data of recognition are saved to improve the learnability of recognition and further ensure the accuracy of recognition.

[0035] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A Chinese painting recognition method based on ink analysis, characterized in that, it includes the following steps: 1) First, establish a basic database, collect data of Chinese paintings from modern times and before through the Internet, and scan some Chinese paintings without existing data on the Internet with a scanning device to obtain sample images and sample data; 2) Then establish a real-time database, place the Chinese painting to be recognized in the scanning device for image acquisition and data acquisition, and store the obtained acquired data inside the real-time database; 3) Verify the data collected in step 1 with professionals, set time-based and character-based classification intervals inside the basic database, and finally store the verified data classified; perform inclusion processing between the time-based and character-based classification intervals, and associate each character-based data to the time-based interval; 4) Establish an overall image grid board for recognizing Chinese paintings, and establish and display a grid template that is evenly divided into several equal parts and suitable for the screen according to the observed image on the display screen system; The grid template realizes that when any grid square of the template is tapped with a finger each time, the effect of the grid is displayed, and the corresponding grid box has no limit on shape and no limit on color of the mark; 5) Transmit the image collected inside the real-time database to the display screen, and output the type of the ink style of the Chinese painting to be recognized after comparing the ink style of the Chinese painting to be recognized with the ink style data in the character-based classification interval collected in the basic database through the image recognition and analysis module; 6) Make a data judgment on the ink data of the Chinese painting to be recognized collected inside the real-time database and the existing ink data of Chinese paintings collected in the basic database, and output the type of the ink of the Chinese painting to be recognized; 7) Overlap the obtained judgment information in steps 5 and 6 in terms of historical relationship to make a final judgment on the recognition of the Chinese painting; Establish a log storage classification interval in the basic database; Record the data of the Chinese painting to be recognized and the recognition steps to generate a log and retain it in the log storage classification interval in the basic database.

2. A Chinese painting recognition method based on ink analysis according to claim 1, characterized in that, the sample data and the data of the Chinese painting to be recognized in steps 1 and 2 include ink spectrum data and multi-light source detection data.

3. A Chinese painting recognition method based on ink analysis according to claim 2, characterized in that, after the ink spectrum data and the multi-light source detection data are processed, one or more characteristic data are obtained.

4. A Chinese painting recognition method based on ink analysis according to claim 3, characterized in that, the characteristic data includes data formed by spectro-spectral data, peak values associated with the spectro-spectral data, fluorescence quantization data, infrared light quantization data, ultraviolet light quantization data, extended light quantization data, color, type, and creation time.