Identification of the morphological defects present in the pattern of spermatozoa using a reconfigurable device

The term “pattern recognition” encompasses a wide range of information processing problems of great practical significance, from speech recognition to medical diagnosis. This paper presents the idea of identification of morphologically defective sperm using pattern recognition and a neural network. Research into determining the infertility rate of sperm is under constant development. First, an important phase in sperm infertility observation is spermatozoon detection. Keeping this in mind, this paper focuses on how to detect the spermatozoon image using an Artificial Neural Network (ANN). For this, “Img capture” software and Visual Studio were employed. Later, Matlab was used to binarize the images and find coordinate points. The Final step was to implement the design using a Field Programmable Gate Array (FPGAs). The Moore-Neighbor algorithm was chosen to trace the edges of the images. Finally, results were obtained from Matlab, showing the image matching percentage per form. This aids in the successful extraction of the spermatozoon shape.