Method for preprocessing data from edge end camera to algorithm reasoning process

Abstract

The invention discloses a method for preprocessing data from an edge end camera to an algorithm reasoning process; the method comprises the following steps: S1, starting data processing, and selecting cameras of different specifications in specific scenes; s11, judging whether the resolution of the camera meets the requirement or not, if the resolution does not meet the requirement, returning to the step S1, and if the resolution meets the requirement, turning to the step S12; and S12, judging whether the power meets the requirement or not, and if the power does not meet the requirement, returning to S1. The invention provides a method for preprocessing data from an edge end camera to an algorithm reasoning process. The method is applicable to industrial fields such as precision agriculture and high-precision data surveying and mapping, and the data quality is managed and controlled in multiple dimensions by adopting an SPV data acquisition method, an MPP data processing method and an LM data labeling method according to specific corresponding scenes. Therefore, the defect that in existing surveying and mapping and precision agriculture scenes, the precision of a deep learning algorithm cannot be further improved in data preprocessing, and consequently landing is difficult is overcome.

Description

technical field [0001] The invention relates to the technical field of data preprocessing, in particular to a data preprocessing method from an edge camera to an algorithm reasoning process. Background technique [0002] With the popularization and commercialization of deep learning, when a large number of deep learning algorithms need to perform model training and learning on multi-scenario data, only by training high-quality image data obtained from the edge can it be more powerful. The accuracy of the same model is maximized. One of the key technologies to achieve high-precision algorithms is to divide and preprocess the scenario-based data for different business types. According to different data division methods and data preprocessing methods, Effective data cleaning of raw data. [0003] The accuracy of deep learning algorithms cannot be further improved for data preprocessing in existing surveying and precision agriculture scenarios, resulting in difficulties in impl...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a method of data preprocessing from the edge camera to the algorithm reasoning process, so as to solve the problem that the accuracy of the deep learning algorithm cannot be further improved in the existing surveying and mapping and precision agriculture scenarios in the above-mentioned background technology. As a result, it is difficult to implement. In scenarios such as surveying and mapping, precision agriculture, etc., the technical reserves of data preprocessing methods applied under deep learning are almost zero. In the absence of a relatively complete data preprocessing solution, deep learning algorithms are applied to the end side It is difficult to achieve high precision and high efficiency in the field of scenarios, which cannot meet the business needs of job scenarios

Images