Computer imaging control method based on image identification

Abstract

The invention discloses a computer imaging control method based on image identification. The computer imaging control method comprises the following steps of firstly, obtaining and processing waveform data, establishing an initializing image, utilizing image identification and optimizing classification to select representative time and space data, performing computer reverse revolution imaging on the optimized data, and outputting a parameter image of an unknown object; building a parameter image model with a fuzzy structure, and setting a threshold value to classify a data set, so as to reduce the data amount; utilizing an image identifying technique to identify an abnormal area with liability of energy storage from a high-dimensional parameter image of the unknown object structure outputted by the reverse revolution imaging; feeding back to judge whether more data of the abnormal area is required for computer imaging again. The computer imaging control method has the advantage that on the basis of image identification and machine learning algorithm, waveform data and image data after computer imaging, the time and space data is subjected to image identification and optimizing classification in the data pre-processing phase; the image data with higher similarity can be interchanged, so as to reduce the data amount.

Description

technical field [0001] The invention belongs to the field of computer imaging, in particular to a computer imaging control method based on image recognition. Background technique [0002] With the enhancement of hardware computing power, computer imaging has been more widely and deeply applied in various practical problems such as energy survey, CT imaging and weather forecast. Due to the huge amount of spatio-temporal data collected by sensors and their related networks, if these data are used as direct input for imaging, although the data can be fully utilized to a certain extent, it cannot guarantee high imaging quality and will also cost At the same time, if the data is not preprocessed, it is easy to cause confusing pixels in the output structure parameter image. Image recognition method is an important tool to solve this series of problems. [0003] At present, almost all energy survey companies, such as: Chevron, PetroChina, BP, ExxonMobil, etc. have used various im...

AI Technical Summary

Problems solved by technology

[0009] The present invention preprocesses the problem that the waveform data volume is large and difficult to process, that is, the technical problem to be solved by the present invention is to provide a computer imaging control method based on image recognition, which can solve the problem that the waveform data volume is large and difficult to quickly calculate and process. and the problem of image sensitive area recognition and detection

Images