Sistema inteligente para reconhecimento de nematodas – Cicla ocellaris em pescada branca-Plagioscion squamosissimus da família Sciaenidae utilizando detecção de objetos em tempo real

Abstract

This work presents a detailed proposal on the detection of parasite points using Deep-learning. The reason why the study is carried out refers to the problem of procedures that are manual and use human vision and, therefore, causing a delay or errors in inspection depending on the analyzed quantity of some parasites in fish. Among the image tools there are options for recognizing patterns that are based on computational applications that can recognize voices, images and the like. As a result, shape recognition and detection analyzes are focused on mathematical calculations involving areas and perimeters, and statistical parts of the patterns chosen to detect larvae in peacock bass. Thus, this work discusses the main functions for pattern recognition with the use of digital processing filters, the results with the number of larvae and detection of areas with them, an image will be generated in IDE’S with Python, OV7670 camera , and the like, to convert the data generated by the camera into machine language or an existing photo of the studied object