A method for concurrent processing of real-time recording on Android platform

Abstract

The invention discloses a method for concurrent processing of real-time recording on an Android platform, which includes a, at the Android HAL layer, after acquiring hardware nodes and collecting audio data, creating a server to be responsible for connection management and data transmission of the receiving client ; b, at the System Framework layer, create a client, connect to the server during initialization, and set the audio data source of the server as the data of the client; c, at the application layer, create a Record SDK, the Record SDK Responsible for initializing the application and connecting with the server to receive data, at the same time maintaining the abnormal connection or application abnormality, and providing an interface for the application to obtain audio data. Using the above method, the application can bypass the Audio architecture of the System Framework, and can support multiple audio applications at the same time, without considering the problem of resource occupation, and without using independent hardware modules to process additional audio data and wake-up mechanisms, saving The extra cost of the product simplifies the technology greatly.

Description

technical field [0001] The invention relates to real-time recording, in particular to a method for concurrent processing of real-time recording on an Android platform. Background technique [0002] At present, smart voice devices based on the Android platform are more and more common in life. Limited by the recording mechanism (recorder) of Android, the initial smart voice devices basically have a single scene, and queue processing for real-time recording: very In the active state, the recording resource is used for continuous monitoring. After the key wake-up word is recognized, the resource is handed over to the recognition engine, and the recording resource is handed over to the monitoring after the recognition is completed. [0003] The current upgrade solution on the market is to mix monitoring and recognition, monopolize real-time recording, and process data internally, without additional resource release and acquisition operations at the system layer. However, when v...

AI Technical Summary

Problems solved by technology

However, when voice and video chatting, karaoke, audio and video recording, etc. need to monopolize the recording resources for a long time in the interactive scene, the voice wake-up and recognition cannot work normally because the recording resources cannot be obtained.

At present, similar products will choose hardware solutions in terms of technical solutions to make up for the defects caused by the mechanism, and the additional audio data and wake-up mechanism will be processed by independent hardware modules, which also causes additional costs and more complex technical challenges for the products.

Images