Forty milligrams of DHB matrix was dissolved in 1 mL of 80:20 (v/v) methanol/water solution containing 0.2% TFA. The matrix solution was mixed with 20 μM ACTH aqueous solution in the ratio 1:1 (v/v). The 200 μL of the mixture was continuously sprayed on a stainless steel MALDI plate using an airbrush pistol GRAFO T1 (Harder & Steenbeck Airbrush, Norderstedt, Germany) with a 0.15 mm orifice from a distance of 8 cm under a working pressure of 300 kPa.
For multicellular spheroid formation, the HT-29 cells were washed with 1 mM EDTA in 1× PBS, mixed with fibroblasts (1:1) and seeded at a density of 50,000 cells/mL on the 12-well plate in DMEM supplemented with 1% FBS and 2 mM L-glutamine. The cellular mixture was incubated on a rotary shaker (60 rpm, Orbital Shaker, NB-101SRC, N-BIOTEK, Korea) in humidified 5% CO2 atmosphere at 37 °C for 1 day. Then, the culture medium was removed and replaced with fresh DMEM medium supplemented with 10% FBS, 2 mM L-glutamine and the spheroids were cultivated for 5 days. The grown spheroids were treated with 20 μM perifosine for 24 h.
In our previous work , we proved the concept of fast MSI using rapid laser beam microscans. Originally, the laser beam was focused onto the sample surface by a lens with a 254-mm focal length at an angle of 20° (referenced to the axis orthogonal to the plate) bypassing both grid electrodes to prevent the laser beam from being blocked by the grids. However, this resulted in the elliptic shape of the laser spot on the MALDI target, which limited the MSI lateral resolution along the horizontal axis. Obviously, a much better lateral resolution can be achieved by decreasing the angle of incidence and aiming the laser beam through the grids, thus producing a circular spot. To further decrease the spot size, a lens with a shorter focal length should be used.
The more pixels acquired in a single linescan with the laser beam, the higher the gain in acquisition rate is as less stage translations between the scan segments (lines) are required. This trend is illustrated in Figure 4, which shows an increase in the acquisition rate with the reduction of the pixel size.
MSI acquisition rate in laser beam scanning and target translation mode utilizing flyback scan pattern as a function of the pixel size, number of laser shots per pixel (pulses), and data points per spectrum (dp). Laser repetition rate was 5 kHz
Overall, the acquisition rate was limited by the slowest of the three repeating events: the spectrum acquisition, pixel-to-pixel translation, and data processing/storage. It has to be noted that certain time was also spent on software control and command sending. This time was not explicitly measured and it is included in the data processing/storage time.
The MSI acquisition rate in the laser beam scanning mode was determined by all three parameters, dependent on the particular measurement conditions. Using 100 μm pixels, the pixel-to-pixel translation was the rate-limiting factor with 30 ms per pixel, approximately. With the reduction of the pixel size to 10 μm, the pixel-to-pixel translation time shortened to approximately 2 ms due to dominant involvement of the optical scanner in pixel addressing. Also, when using 100 laser shots per pixel, the acquisition rate was limited to 40 pixel/s by spectra acquisition (20 ms per pixel at 5 kHz). Going down to 10 shots per pixel, the acquisition rate further rose to values over 100 pixel/s and the data processing/storage became a major consideration with approximately 6 ms per pixel. Finally, storing 40,000 instead of 100,000 data points per pixel increased the acquisition rate to 147 pixel/s. Still, the data processing/storage rate was the major limiting factor, adding approximately 3 ms per pixel. This time constraint is expected to gradually diminish due to the constantly increasing speed of processors.
In summary, utilizing the 5-kHz laser and 1-mm-long laser beam scans provided an acquisition rate of 147 pixel/s for 10-μm raster with 10 pulses fired per pixel and 40,000 data points in the spectra, see Figure 4. This performance surpasses the maximal published acquisition rate 50 pixel/s of the above mentioned commercial MALDI TOF MS instrument with two optical scanners . MSI acquisition rates of 100 pixel/s were so far reported only for inductively coupled plasma (ICP) TOF MS system for elemental mapping of the surfaces .
Multiple-license plate recognition is gaining popularity in the Intelligent Transport System (ITS) applications for security monitoring and surveillance. Advancements in acquisition devices have increased the availability of high definition (HD) images, which can capture images of multiple vehicles. Since license plate (LP) occupies a relatively small portion of an image, therefore, detection of LP in an image is considered a challenging task. Moreover, the overall performance deteriorates when the aforementioned factor combines with varying illumination conditions, such as night, dusk, and rainy. As it is difficult to locate a small object in an entire image, this paper proposes a two-step approach for plate localization in challenging conditions. In the first step, the Faster-Region-based Convolutional Neural Network algorithm (Faster R-CNN) is used to detect all the vehicles in an image, which results in scaled information to locate plates. In the second step, morphological operations are employed to reduce non-plate regions. Meanwhile, geometric properties are used to localize plates in the HSI color space. This approach increases accuracy and reduces processing time. For character recognition, the look-up table (LUT) classifier using adaptive boosting with modified census transform (MCT) as a feature extractor is used. Both proposed plate detection and character recognition methods have significantly outperformed conventional approaches in terms of precision and recall for multiple plate recognition.
With the ever-increasing traffic situations in modern cities, the demand for the ITS is also increasing rapidly. License Plate Recognition (LPR) is a crucial component of an ITS, which is used to identify vehicles based on their number plates. The LPR includes license plate localization to find the location of the plate in an image, followed by segmentation and recognition of alphanumeric characters of the localized plate. Most of the LPR methods are capable of recognizing a single vehicle in an image .
In recent years, there is a considerable increase in problems of traffic congestion, security monitoring, and over-speeding in modern cities . The afore-described phenomenon has increased the demand for identifying multiple vehicles in an image. A high-resolution camera can monitor multiple lanes containing several vehicles. However, recognizing multiple vehicles becomes challenging as some plates will have a smaller size or low resolution, for instance, based on distance from the camera, different background colors, distortions, and different contrast as shown in Fig. 1. Moreover, the HD images also increase computational cost .
Several methods have been proposed for license plate recognition over the years, such as template matching , artificial neural networks , adaptive boosting , and Support Vector Machines (SVM) . Most of these methods performed well in constrained environments, such as uniform illumination and fixed plates size. In the recent past, Deep Learning (DL) methods have been widely used as powerful tools for image recognition applications . Henceforth researchers have also developed various Convolutional Neural Network (CNN) architectures for plate recognition . Cascaded CNN  has a high computational cost and multi-CNN  struggles on varying parameters of size and angle. Faster-RCNN has high speed but performance deteriorates on low-resolution plates and non-uniform illuminations . Therefore, it can be concluded that existing approaches have yet to achieve a satisfactory level when exposed to unconstrained environments, such as varying illuminations conditions, colored background plates, and variations in plate, and font sizes. Therefore, this paper proposes a novel MLPR system that is capable to handle all the aforementioned problems. The major contributions of this paper are as follows:
The proposed technique improves the accuracy of plate detection in challenging environments that have non-uniform illumination and low resolution (based on distance from the camera). Our proposed MLPR technique divides plate detection problems into vehicle detection and plate localization, which results in scaled information for plate localization and helps to remove background noise and clutters.
In this paper, Faster-RCNN is used for vehicle detection followed by plate localization using morphological operations in the HSI color space. Geometric properties of area and aspect ratio of connected pixels are used for character segmentation. Moreover, this paper uses texture-based feature extraction method MCT, which is robust to illumination changes and low resolution , with lookup table classifier in boosting framework for character recognition.
The remainder of this paper is organized as follows. Related work is reviewed in Sect. 2. Section 3 presents the proposed license plate detection and recognition method, detailed simulation results are presented in Sect. 4. Finally, the conclusion is given in Sect. 5.
Most of the existing work on plate detection target a single vehicle in an image. Therefore, the demand for multiple plate detection has increased considerably owing to an increase in multilane structure in modern cities. Edge detection methods consider an area with a higher density of characters as an LP. Combining this property with geometric properties of plates has been widely used to extract LPs. Vertical edge detection is more robust compared with horizontal edge detection, which provides inaccurate results owing to errors due to the car bumper area . A fast and robust vertical edge detection method was proposed that increases the speed of detection by eliminating unwanted lines . Yepez et al.  proposed a plate detection method based on only morphological operations. They developed an algorithm to select appropriate structuring element (SE) from a set of SEs by training these SEs on the whole dataset. This approach could not perform well for multiple license plate recognition, due to variations in the size of plates in an image. 2b1af7f3a8