Development and preliminary validation of an automatic digital analysis system for animal fibre analysis
2017
Quispe, M. D.(MAXCORP Technologies) | Benavidez, G.(Universidad Nacional de Trujillo Escuela Académico Profesional de Zootecnia) | Sauri, R. A.(MAXCORP Technologies) | Bengoechea, J. J.(MAXCORP Technologies) | Quispe, E. C.(National University Autonoma de Chota)
New equipment was designed and developed to evaluate the physical characteristics of animal fibres, based on an automatic digital analysis system (ADAS) that allows the capture of a series of real-time images. In the development of the mechanical component, a design tool was used that allows visualization, simulation, and documentation of the product. At the same time, an Atmel (ATmega328) microcontroller was programmed to enable displacement of table coordinates, focus of images, and reading of the temperature and humidity of fibre samples and the environment. The fibre images were processed using artificial vision technology. Algorithms were also developed for edge detection to define the diameter of the fibre in pixels. Finally, calibration was carried out using a regression and standardized samples of wool tops. The authors then weighted the pixels to µm with a standard sample. A friendly graphical interface was developed for management of the built equipment, visualization of results, calibration, data and graphic export, configuration, among others. For validation, average fibre diameter (AFD), standard deviation (SD), coefficient of variation and comfort factor (CF) were compared with values of top wool patterns. Finally, the measurements of OFDA 2000 and FIBER-EC were compared using student t-test and Pearson correlation. The results of the validation showed that the confidence limit of FIBER-EC, which varied between 0.075 and 3.47 µm, is similar to that of the confidence limit of Sirolan Laserscan and OFDA 2000. Accuracy is better than the OFDA 2000 for fibre assessments less than 25 µm, which vary between 0.034 and 0.250 µm.
显示更多 [+] 显示较少 [-]