Mechanical Engineering of Biosystems Department, Razi University, Kermanshah, Iran.
Environ Sci Pollut Res Int. 2021 Jul;28(26):34501-34510. doi: 10.1007/s11356-021-12974-y. Epub 2021 Mar 2.
In this research, a new android app for smartphones for estimating some water quality parameters in carp fish ponds such as pH, electrical conductivity (EC), total dissolved solids (TDS), and turbidity is presented. Contact imaging was used to acquire images from the samples. To estimate pH, EC, TDS, and turbidity values, 12 features were extracted from each image. Features were used as input to the artificial neural network models. The performance of the models was evaluated by the R and RMSE parameters. Based on the results, the network with a structure of 12-15-4 was selected as the best model. The values of R for estimating pH, TDS, EC, and turbidity were 0.913, 0.993, 0.994, and 0.958, respectively, while the corresponding values for the RMSE were 0.054, 1.835, 3.766, and 0.262, respectively. Finally, this model was successfully implemented on an app named WaterApp on the android smartphone. For testing the app on the smartphone, the performance of the model was evaluated again using new images. According to the results, the R values for validation data by the developed WaterApp for pH, EC, TDS, and turbidity were 0.88, 0.913, 0.884, and 0.944, respectively.
在这项研究中，我们开发了一款新的安卓智能手机应用程序 WaterApp，用于估算鲤鱼养殖池塘中的一些水质参数，如 pH 值、电导率（EC）、总溶解固体（TDS）和浊度。该应用程序使用接触成像技术从样本中获取图像。为了估算 pH 值、EC、TDS 和浊度值，从每个图像中提取了 12 个特征。这些特征被用作人工神经网络模型的输入。通过 R 和 RMSE 参数评估模型的性能。结果表明，结构为 12-15-4 的网络被选为最佳模型。用于估算 pH 值、TDS、EC 和浊度的 R 值分别为 0.913、0.993、0.994 和 0.958，而相应的 RMSE 值分别为 0.054、1.835、3.766 和 0.262。最后，该模型成功地在安卓智能手机上的 WaterApp 应用程序上实现。为了在智能手机上测试该应用程序，我们再次使用新图像评估模型的性能。结果表明，开发的 WaterApp 应用程序对 pH 值、EC、TDS 和浊度验证数据的 R 值分别为 0.88、0.913、0.884 和 0.944。
