Method of Determining Characteristic of Image-Based Sensors Using Compact yet Efficient Cascade Context-based Estimation Model Algorithm

Obtaining a continuous relationship between input and output is essential during sensor characteristic determination. However, it is a little challenging for complicated input data, for example, image data. In this article, a sensor characteristic determining method using deep learning Compact yet Efficient Cascade Context-based Estimation (C3E) network, which can establish the continuous relationship between the input and the output, was proposed. A UV sensing system based on this C3E method was presented. The sensor after algorithm optimization has a sensitivity of $1~\mu \text{W}$ /cm2. The limit of detection is $9~\mu \text{W}$ /cm2. The network test time is close to 3 s, and the error rate is 5%, which is very suitable to transplant to the embedded platform and provides solutions for compact instruments.