The Number Of Silkworm Eggs Statistical Programs Based On Image Processing

We propose here a method to count the silkworm eggs automatically based on the image processing technology, which differs from traditional manual counting method and avoids such defects as "eye strain, low efficiency, lack of accuracy, easy to damage small silkworm eggs" etc.We first study the image processing algorithms including color conversion, binarization, error correction which are needed in the preprocess of the silkworm eggs image. For the great importance of image segmentation to the counting work, we analyze in depth the corrosion morphology and watershed transform algorithms which has been improved by us afterwards. Based on the above efforts, we put forward the best image processing solutions by comparing the experimental results, which improves the statistical accuracy significantly. Thus the real-time and efficient measurement has been realized. In addition, manual counting is taken as the last step to guarantee the correct rate of 100%.We write the application program in C++ language, in which Windows WIA technology is applied to collect images of silkworm eggs and the above improved algorithms are used to accomplish the image processing and counting work. The experimental results show that the correct rate of automatic counting is above 97%, which reaches a satisfactory result. Accompanied by manual calculation, the number will be completely correct. The whole process takes only a few seconds.