Drosophila Compound Eye Lesion Detection In Machine Vision Systems

To speed up the development of biomedicine has been the goal of the scientific researchers for many years .With technical development, scientific researchers research it further and further. The plentiful types of the drosophila's can provide favorable experimental animal models for scientists who study on the genetics and the developmental biology. This paper studies on the theories and technologies of drosophila's compound eyes lesion discrimination based on the machine vision system. It is a novel attempt to make use of the machine vision technology to study the kind of biomedical micro-image domain such as drosophila and mini-type samples. This work will be worth studying.It has made some effective improvements in the following five aspects: the liquid flow drosophila samples transportation, drosophila's compound eyes micro-image collection, drosophila's compound eyes lesion determination, drosophila's compound eyes lesion recognition, the diagnosis pictorial information management. Based on the machine vision technology, using quite advanced JVC/TK-C721EC colored camera, stereoscopic biology optical microscope, NI image collection card PCI-1409 and so on, organize and complete the drosophila's compound eyes micro-image collections. It settles the base for drosophila's compound eyes lesion tests. Based on image processing software Vision Assistant, it realizes three kind of extraction the contour line of the drosophila's compound eyes algorithm-grayscale transformation, close operation, color threshold, comparing and analyzing respective applicable scope and function. It also indicates that the extraction method of the close operation is superior to other methods .The designed system has realized the quick analysis, the computation and provided results of the drosophila's compound eyes bad area position and the bad spot size.Finally, based on the LabVIEW software as the platform, human-computer interface for automatic detection of pathological changes of drosophila's compound eye was established. The real-time process of image processing and related results such as the area of compound eyes and bad spot within it was displayed, the choice of the contour line or bad spot of the drosophila's compound eyes ,threshold setting, image pre-processing module, image parameter, close operation and color threshold module .Compared with manual inspection, the detection precision of this system has achieved 90% and the processing time has been 1.5-1.8s/sheet. These will be a new effective method to test the drug reaction of the subject (such as drosophila) participated the experiment.