Multimodal image and spectral feature learning for efficient analysis of water-suspended particles.

We have developed a method to combine morphological and chemical information for the accurate identification of different particle types using optical measurement techniques that require no sample preparation. A combined holographic imaging and Raman spectroscopy setup is used to gather data from six different types of marine particles suspended in a large volume of seawater. Unsupervised feature learning is performed on the images and the spectral data using convolutional and single-layer autoencoders. The learned features are combined, where we demonstrate that non-linear dimensional reduction of the combined multimodal features can achieve a high clustering macro F1 score of 0.88, compared to a maximum of 0.61 when only image or spectral features are used. The method can be applied to long-term monitoring of particles in the ocean without the need for sample collection. In addition, it can be applied to data from different types of sensor measurements without significant modifications.

Guiding Labelling Effort for Efficient Learning With Georeferenced Images.

We describe a novel semi-supervised learning method that reduces the labelling effort needed to train convolutional neural networks (CNNs) when processing georeferenced imagery. This allows deep learning CNNs to be trained on a per-dataset basis, which is useful in domains where there is limited learning transferability across datasets. The method identifies representative subsets of images from an unlabelled dataset based on the latent representation of a location guided autoencoder. We assess the method's sensitivities to design options using four different ground-truthed datasets of georeferenced environmental monitoring images, where these include various scenes in aerial and seafloor imagery. Efficiency gains are achieved for all the aerial and seafloor image datasets analysed in our experiments, demonstrating the benefit of the method across application domains. Compared to CNNs of the same architecture trained using conventional transfer and active learning, the method achieves equivalent accuracy with an order of magnitude fewer annotations, and 85 % of the accuracy of CNNs trained conventionally with approximately 10,000 human annotations using just 40 prioritised annotations. The biggest gains in efficiency are seen in datasets with unbalanced class distributions and rare classes that have a relatively small number of observations.

Unsupervised feature learning and clustering of particles imaged in raw holograms using an autoencoder.

Digital holography is a useful tool to image microscopic particles. Reconstructed holograms give high-resolution shape information that can be used to identify the types of particles. However, the process of reconstructing holograms is computationally intensive and cannot easily keep up with the rate of data acquisition on low-power sensor platforms. In this work, we explore the possibility of performing object clustering on holograms that have not been reconstructed, i.e., images of raw interference patterns, using the latent representations of a deep-learning autoencoder and a self-organizing mapping network in a fully unsupervised manner. We demonstrate this concept on synthetically generated holograms of different shapes, where clustering of raw holograms achieves an accuracy of 94.4%. This is comparable to the 97.4% accuracy achieved using the reconstructed holograms of the same targets. Directly clustering raw holograms takes less than 0.1 s per image using a low-power CPU board. This represents a three-order of magnitude reduction in processing time compared to clustering of reconstructed holograms and makes it possible to interpret targets in real time on low-power sensor platforms. Experiments on real holograms demonstrate significant gains in clustering accuracy through the use of synthetic holograms to train models. Clustering accuracy increased from 47.1% when the models were trained only on the real raw holograms, to 64.1% when the models were entirely trained on the synthetic raw holograms, and further increased to 75.9% when models were trained on the both synthetic and real datasets using transfer learning. These results are broadly comparable to those achieved when reconstructed holograms are used, where the highest accuracy of 70% achieved when clustering raw holograms outperforms the highest accuracy achieved when clustering reconstructed holograms by a significant margin for our datasets.

Identification of microplastics in a large water volume by integrated holography and Raman spectroscopy.

A noncontact method to identify sparsely distributed plastic pellets is proposed by integrating holography and Raman spectroscopy in this study. Polystyrene and poly(methyl methacrylate) resin pellets with a size of 3 mm located in a 20 cm water channel were illuminated using a collimated continuous wave laser beam with a diameter of 4 mm and wavelength of 785 nm. The same laser beam was used to take a holographic image and Raman spectrum of a pellet to identify the shape, size, and composition of material. Using the compact system, the morphological and chemical analysis of pellets in a large volume of water was performed. The reported method demonstrates the potential for noncontact continuous in situ monitoring of microplastics in water without collection and separation.

Partial least squares regression calculation for quantitative analysis of metals submerged in water measured using laser-induced breakdown spectroscopy.

Effects of different parameters regarding partial least squares (PLS) regression analysis are investigated for quantitative analysis of water-submerged brass samples. The concentrations of Cu and Zn in various brass alloys were quantified using PLS, and the performance after different signal processing steps (normalization, smoothing, and background subtraction) and database segmentation by excitation temperature is compared. In addition, the effects of averaging numbers on the results are examined. From the results, normalization was found to be the most effective among three established signal processing methods. The effects of both peak and background fluctuations seen in the signals are reduced by normalization. It was found that temperature segmentation of the database in an appropriate range, which should be high enough for reliable peak detection, can further improve the accuracy of PLS calculations. The proposed method is applicable in real time, and can potentially be used for automated fast and accurate measurements of solids at oceanic pressures.

On-site quantitative elemental analysis of metal ions in aqueous solutions by underwater laser-induced breakdown spectroscopy combined with electrodeposition under controlled potential.

We propose a technique of on-site quantitative analysis of Zn(2+) in aqueous solution based on the combination of electrodeposition for preconcentration of Zn onto a Cu electrode and successive underwater laser-induced breakdown spectroscopy (underwater LIBS) of the electrode surface under electrochemically controlled potential. Zinc emission lines are observed with the present technique for a Zn(2+) concentration of 5 ppm. It is roughly estimated that the overall sensitivity over 10 000 times higher is achieved by the preconcentration. Although underwater LIBS suffers from the spectral deformation due to the dense plasma confined in water and also from serious shot-to-shot fluctuations, a linear calibration curve with a coefficient of determination R(2) of 0.974 is obtained in the range of 5-50 ppm.

Distribution of local 137Cs anomalies on the seafloor near the Fukushima Dai-ichi Nuclear Power Plant.

An estimated 3.5±0.7×10(15) Bq of (137)Cs is thought to have been discharged into the ocean following the melt down at Fukushima Dai-ichi Nuclear Power Plant (F1NPP). While efforts have been made to monitor seafloor radiation levels, the sampling techniques used cannot capture the continuous distribution of radionuclides. In this work, we apply in situ measurement techniques using a towed gamma ray spectrometer to map the continuous distribution of (137)Cs on the seafloor within 20 km of the F1NPP. The results reveal the existence of local (137)Cs anomalies, with levels of (137)Cs an order of magnitude higher than the surrounding seafloors. The sizes of the anomalies mapped in this work range from a few meters to a few hundreds of meters in length, and it is demonstrated that the distribution of these anomalies is strongly influenced by meter scale features of the terrain.