Note (General)Quality evaluation of koi (Cyprinus rubrofuscus) is essential to koi industry. However, the community has a lack of knowledge about the quality evaluation of koi and about features of koi which are valuable. During the koi exhibitions like "All Japan Nishikigoi Show", quality evaluation of koi is conducting by authoritative experts. And the information about the evaluation methods used is not known to the community. While koi quality rate is the cornerstone for this ornamental fish. In this study, photographs of participants in the koi exhibitions were used for analysis to discover features that affect the quality of koi varieties of Kohaku. The HSVA color model was used to define and extract from image red and pale red coverage rates in Kohaku body coloration. Body aspect ratio, body proportion values, red coverage ratio, and pale red coverage ratio were extracted using Python programming language tools and analyzed using the R statistical software. Results of analysis of extracted data show that significant impact on Kohaku quality have: the body aspect ratio (p < 0.001) and the pale red coverage rate (p < 0.001). For the body aspect ratio, there was a -0.236 (±0.105) decrease of the quality rate for each extra unit of aspect ratio. For each extra percent of pale red coverage rate, the quality rate decreases by -0.037 (±0.004). According to these results pale red coverage rate is more important feature than body aspect ratio. In the process of the present research, we encountered such problems as inappropriate koi body posture on the photo and poor image quality. An inappropriate koi body posture is one in which the koi's body is bent and/or carp has fins in an asymmetrical position. The possible reasons of poor image quality are poor lighting, water reflection, indirect camera position, low photo resolution or fuzzy image. We solved these problems with selecting images with appropriate image quality and koi posturer. Also, a serious problem was that the position of the fins varies from image to image, making it impossible to compare them. This problem was solved by cropping the images. All these problems make analysis difficult, leading to the need for quality control of materials. The number of materials is limited, which leads to the need for self-collection of materials. In addition, the type of material itself, namely photographs, is a limiting factor, as photographs provide a limited representation of a koi's body structure in three-dimensional space. This leads to limitations of the methods that can be used for analysis. Thus, this prompted us to develop a method of collecting materials that would allow us to gain a more complete information of the koi body and, in turn, evaluate its quality. The advantages and disadvantages of the materials and methods used in the study were discovered and analyzed. Based on which new ways of collecting data were invented for the further development of this topic. The most promising method for collecting materials, in our opinion, is the use of a neural network to automate the interpolation of the shape of a 3D koi template model and texturing. The present study and its results are an important foundation for the further research into the quality of koi and for the development of the koi industry in general.
新大院博(農)第231号
Collection (particular)国立国会図書館デジタルコレクション > デジタル化資料 > 博士論文
Date Accepted (W3CDTF)2023-05-05T22:12:34+09:00
Data Provider (Database)国立国会図書館 : 国立国会図書館デジタルコレクション