Key points are not available for this paper at this time.
The automated identification of colors and their intensity from sensor images is a significant interest in field deployable and cost-effective smartphone-based water monitoring solutions. Artificial Intelligence (AI) has been extensively used in automated image processing applications specifically when there is no recognizable pattern in the image data. AI considerably outperforms conventional detection techniques using image analysis in such scenarios. In the present work, we have developed an Artificial Intelligence (AI) based mobile application platform, that can capture the sensor image using an inbuilt smartphone camera, identify the presence of sensing parameters and classify the level of the same based on color intensity recognized in the training sets of the captured image using deep convolutional neural networks (CNN). As a test case, we have implemented the developed AI-based mobile application platform to monitor the water quality for bacterial contamination where the sensor images are classified into the presence or absence of bacteria based on visual appearance. Our method is seen to detect the presence with a ∼99.99% accuracy which is an improvement in the detection accuracy of the already established method in this regard where manual visual inspection is carried out to classify the sensor images. The considerable enhancement in detection accuracy can be attributed to the elimination of subjective decision making which inevitable consequence of human intervention in the reported test case.
Gunda et al. (Tue,) studied this question.