Studies On Preliminary Intelligent Examination Of Helminth Eggs In Cattle Feces

Ruminant parasitic disease is one of the widespread and harmful disease. The examination of helminth eggs in feces is commonly used method to diagnose and investigate ruminant helminthiasis at the present. This method was most classical and widely used diagnostic method for animal parasitic diseases, but the method is not accurate enough and inefficient, as well as, the examination result was strongly depended on the technicians. Therefore, we attempt to develop an intelligent diagnosis system for animal helminthiasis base on fecal egg examination method and invariant moment.Firstly, cattle gastrointestinal parasite disease was investigated applying Veterinary Epidemiological method in XiangHuangQi area. For establishing image database of helminth eggs, different kinds of images of helminth eggs were collected in the investigation. Furthermore, abundant eggs were selected from identified helminth cloaca through microdissection technique and photographed by biological microscope. Then, the Pseudo-Jacobi-Fourier Moments were used for feature extraction, and egg images reconstruction experiments have been done. Finally, we attempted to realize the interface of egg identification and count system with the graphic user interface (GUI) of Matlab software. The results reveal that:the prevalence of helminthes was44.92%in Inner Mongolia XiangHuangQi area. According to the preliminary identification of morphological characteristics of eggs,9kinds of species were identified attributed to3phyla,3classes,3orders,5families and9genera, respectively.9kinds of original image database of helminth eggs had been established, it contains a total of1147images. The invariant moment database had been established base on the image database. The invariant moment values of the same kind of egg were almost the same, but different kinds of eggs revealed large difference in the invariant moment values through the feature extraction. The interface of eggs identification and count system was preliminary realized, and it ran normally. So, PJFM’s can be used for feature extraction and provide important technology basis for intelligent identification system.