Recognition of fish categories using deep learning

Abstract: The two most critical components of fisheries nowadays are fish classification and fish location. Due to problems with segmentation, noise, and shifting environmental circumstances, it is challenging to categorise the images accurately. However, there is a big market for more precise object recognition. Neural Convolutional Network,VGG16, It enhances the accuracy of classifying and locating the photos of 97 percent, is utilised in order to tackle this challenge. The dataset was used to train a fish classification algorithm, and many activation functions have improved accuracy. Finally, after numerous comparisons, a better strategy has been discovered. The accuracy has improved. The successful creation of class labels and region recommendations is aided by localization.. Lastly, localisation and classification of images with better accuracy.

Keywords: Image processing, Fish Category detection, CNN, VGG16.