Particle detection and size recognition based on defocused particle images: a comparison of a deterministic algorithm and a deep neural network

Abstract

The systematic manipulation of components of multimodal particle solutions is a key for the design of modern industrial products and pharmaceuticals with highly customized properties. In order to optimize innovative particle separation devices on microfluidic scales, a particle size recognition with simultaneous volumetric position determination is essential. In the present study, the astigmatism particle tracking velocimetry is extended by a deterministic algorithm and a deep neural network (DNN) to include size classification of particles of multimodal size distribution. Without any adaptation of the existing measurement setup, a reliable classification of bimodal particle solutions in the size range of $$1.14~\upmu \hbox {m}$$ 1.14 μ m –$$5.03~\upmu \hbox {m}$$ 5.03 μ m is demonstrated with a precision of up to 99.9 %. Concurrently, the high detection rate of the particles, suspended in a laminar fluid flow, is quantified by a recall of 99.0 %. By extracting particle images from the experimentally acquired images and placing them on a synthetic background, semi-synthetic images with consistent ground truth are generated. These contain labeled overlapping particle images that are correctly detected and classified by the DNN. The study is complemented by employing the presented algorithms for simultaneous size recognition of up to four particle species with a particle diameter in between $$1.14~\upmu \hbox {m}$$ 1.14 μ m and $$5.03~\upmu \hbox {m}$$ 5.03 μ m . With the very high precision of up to 99.3 % at a recall of 94.8 %, the applicability to classify multimodal particle mixtures even in dense solutions is confirmed. The present contribution thus paves the way for quantitative evaluation of microfluidic separation and mixing processes.

A deep neural network architecture for reliable 3D position and size determination for Lagrangian particle tracking using a single camera

“…In a second step, a local threshold for each depth position was adaptively determined until 90% of the particle images at a certain depth position were declared as valid. For details, see [8]. A final validation step considered the intensity of the particle image [24].…”

“…However, this is always only achieved with additional efforts, either by an increased manufacturing cost of the microchannels or additional equipment for the measurement [2][3][4][5][6]. In recent years, the measurement methods were extended to allow a classification of the particles based on their particle image from flow measurements [7,8].…”

“…However, one surprising result was the fact that using particles of multiple different sizes did not increase the uncertainty in the depth position although a narrow monomodal size distribution is one assumption for classic APTV [13]. A classification with DNNs was carried out in a recent study by Sachs et al [8], where the DNNs outperformed the classic methods in terms of size determination.…”

Inferring the temperature from planar velocity measurements in Rayleigh-Bénard convection by deep learning

“…A suitable tool to partly replace the need to simultaneously measure the temperature may be found in machine learning algorithms. Machine learning emerges as a powerful tool in the field of fluid mechanics [21,22] with more and more applications in, e.g., flow measurement techniques [23][24][25][26], data reduction [27], forecast [28][29][30] and super-resolution [31,32]. Especially deep neural networks turned out to be a powerful tool, and effort is spent to make their predictions consistent with physical laws by incorporating the governing equations, making them "physics-informed" [33][34][35].…”

3D positioning and autofocus of the particle field based on the depth-from-defocus method and the deep networks

“…Dreisbach et al have recently utilized neural networks to enhance the rate of particle detection and reduce the occurrence of false positives, surpassing the capabilities of traditional detection algorithms [16]. Sachs et al have presented deterministic algorithms and deep neural networks that can recognize the size of up to four particle species simultaneously, with a particle diameter ranging from 1.14 µm to 5.03 µm [17]. Leroy et al used both the soft-assignment encoding and the DfD method to determine the intermediate depth for a single object from defocus blur images [18].…”