Improvement in Detection of Chicken Egg Fertility Using Image Processing Techniques

The objective of this paper is to improve the efficiency in predicting the fertility of chicken eggs using digitalised techniques. The traditional technique finds incongruence in predicting the fertility by manual method. By using this system the digital images obtained as the result of candling process is used for fertility prediction. The obtained image is converted to RGB components as the components clearly depicts the inner view of the image than the original one. The mined component is transformed to binary image to determine the yolk dimensions (height x diameter). This dimensions help in the prediction of air room region which helps hatching. The yolk size value that lies close to the range of threshold value is considered fertile. The investigational study results nearly 80 to 85 % of efficiency than the existing techniques. With the same data set, efficiency is compared with manual method and trained neural network algorithm to provide better accuracy.