Deep learning and sound signal-based liquid identification system

Abstract

The invention discloses a deep learning and sound signal-based liquid identification system, which at least comprises a liquid to be detected and a container, and further comprises acoustic sensing equipment, a wireless transmission system and back-end service equipment; the acoustic sensing equipment is used for acquiring at least loudspeaker equipment, sound recording equipment and acoustic analog-to-digital conversion equipment; the wireless transmission system is used for providing transmission service between the data acquisition equipment and the back-end service equipment to realize wireless sensing; and the rear-end service equipment is used for supporting a liquid classification and identification module operation environment, acquiring data from the acoustic sensing equipment, and performing data processing by utilizing liquid acoustic propagation characteristics to complete liquid identification. The liquid identification system is low in cost, safe, reliable and portable, and more than 20 kinds of liquid can be distinguished under the condition that the position is relatively fixed.

Description

technical field [0001] The invention belongs to the technical field of wireless sensory liquid material identification, in particular to a liquid identification method and system based on deep learning (Deep Learning, hereinafter referred to as DL) and sound signals. Background technique [0002] Fluid testing has attracted high research attention in recent years because it can be used to detect water contamination, adulteration, food additives in beverages, and to detect kidney disease using urinalysis. [0003] Existing liquid identification methods utilize optical and electromagnetic signals. Although this type of method has high accuracy, some optical signals such as X-Ray have an impact on human health, and the cost of generating infrared rays in specific bands is high. Electromagnetic signals such as RFID identification technology require additional tags and use probes for identification. Usually, the analysis equipment is large in size, not easy to carry, and will in...

AI Technical Summary

Problems solved by technology

However, it is still difficult to identify materials using acoustic signals. First, there are many types of liquids, and the differences between many liquids are very small.

Second, it is difficult to extract non-contact features that can be used for identification in the echo signal, and such features are often affected by environmental noise and multipath

Third, container restrictions, special professional equipment and special professionals are not easy to obtain

The main disadvantage of this method is that it requires special equipment and cannot achieve ubiquitous computing and large-scale wide application.

[0008] In summary, the existing methods mainly have the following three types of problems: 1. The detection effect is not good; 2. Professional equipment and personnel are required; 3. It is not easy to be widely used; 4. The safety cannot be guaranteed

These problems lead to existing liquid identification methods with more restrictions, inconvenience and new safety challenges

Images