Intelligent Recognition Translation Optimization Algorithm Based on Machine Vision

Abstract

At this stage, machine vision (MV) technology is developing rapidly, and good results have been achieved in the fields of text recognition, image processing, and natural language processing. Intelligent recognition translation, as an efficient tool, can perform equivalent conversions between different languages, while still maintaining the original semantics, which has important practical significance. Based on this, the purpose of this article is to study the MV - based intelligent recognition translation optimization algorithm. This article proposes a text recognition method based entirely on convolutional neural networks, incorporates attention mechanism, introduces LSTM with "forgetting" function, and introduces the specific structure of a fully convolutional sequence decoder. This article designs an intelligent recognition and translation system, and introduces the end-to-end recognition and translation module. Experimental data shows that the BLUE values of the method in this paper are 25.57, 25.63, and 25.71, respectively. This shows that the method in this paper has a good effect in text recognition and translation.