Study On Detection Method Of Germ Rice Processing Quality

Rice is eaten as a staple food in most parts of the country.Rice is produced after a series of processing such as rice shelling,glutinous rice,and whitening.Now,according to the different processing grades of rice,rice can be divided into the following three categories,namely brown rice,polished rice,and germ rice.Germ rice refers to the ability of the germ to retain more than 80% of the moderately processed product of the rice during the milling process.The germ rice combines the advantages of brown rice and polished rice to form a perfect balance between taste and nutritional value.Under the premise that the domestic living standards are getting higher and higher,people's demand for food is no longer simply to eat,but to eat healthy and eat well.The advantages of germ rice balance the needs of the staple food.At present,the detection of the quality of germ rice mainly depends on the human eye,subjectivity and low efficiency.The research methods for the detection of germ rice quality at home and abroad are single,and the recognition rate is low.In this paper,the image of germ rice is taken as the research object,and the image processing,analysis technology,deep learning technology and fuzzy control technology are applied to detect the quality parameters of germ rice including processing precision and embryo integrity,and the above two indicators are used as input.The pressure value of the grinding mechanism is the output,and a fuzzy control system is designed.The output from the fuzzy control system serves as a feedback parameter to guide the production of germ rice.This paper mainly carried out the following research:First,research on germ rice image acquisition and preprocessing and segmentation techniques was conducted.According to the quality requirements of the image of the germ rice,a set of image acquisition devices is designed,and each part is selected as required.In order to facilitate the subsequent extraction of single rice grains from germ rice,an appropriate filtering algorithm was selected to perform image preprocessing,and a series of filtering smoothing processes were completed,so that the processed images could reach the standard of image segmentation.The image is then segmented to obtain a single germ rice image.Then,the detection of the integrity of the germ and the processing precision of the germ rice is realized,and the acquisition of the feedback parameters guiding the production of the germ rice is studied.The texture feature parameters were extracted based on image texture feature description,and the germline rice processing precision grade classification was performed by svm.The method based on deep learning convolutional neural network was used to classify the embryo integrity.Algorithm;based on the information provided by experienced quality inspectors,design a fuzzy control system that uses the processing precision of germ rice and the integrity of the embryo as the input,the pressure value of the grinding mechanism is the output,and the output is used as the feedback parameter to realize the germ rice.Control of quality processing.Finally,the software and hardware implementation of the germ rice quality detection system is designed.The hardware scheme for designing germ quality detection makes the hardware execution process of germ rice more simple.Automatic image detection and manual detection programs are written based on Matlab2016 and Python 3.6.5.Design the germ rice data cloud control system to realize remote acquisition and control of various information in the process of germ rice processing.