To make High Dynamic Range image sensors easier to integrate in cameras and subsystems, Pyxalis proposes a novel digital architecture, making the management of image acquisition and the reconstruction of HDR images much simpler. This work describes the development and characterization of a 120dB High Dynamic Range sensor based on a novel processor based architecture. Two processors are embedded in the detector to allow flexible shutter operation, windowing, sequencing as well as High Dynamic Range modes. Thanks to built in HDR reconstruction , the image sensor outputs 20bits per pixels based on 14bits column ADCs and two complementary HDR image acquisition methods. The entire image acquisition is fully synchronous.

Spectral imaging has been proved as a promising technology to perform scientific documentation and analysis of cultural heritage objects. Imaging systems used for this purpose varies in different forms and complexity. Most of the spectral devices are designed not specifically for cultural heritage imaging, but for other application domains like remote sensing, satellite imaging etc. While these imaging systems used for cultural heritage scanning, the same processing workflow may not be adequate to meet the required image quality. In this paper, we investigate one of the several quality parameters, geometrical distortions in a spectral image caused by the translation stage of the camera. For this study, we used a hyperspectral image dataset derived from a pushbroom hyperspectral imaging system attached in a rotational translator. Using geometrical model, we have estimated the distortion, which is function of the scanning angle and distance. Image correction has been proposed and tested over a number of images acquired at the laboratory and in a museum environment.

We present a novel single-chip color image sensor with a layered structure of three organic photoconductive films (OPFs), each one sensitive to only one primary color (red, green, or blue). First, we fabricate a red/green component consisting of two OPFs sensitive to red and green colors and transparent readout circuits on a glass substrate for red and green color imaging. Then, we fabricate a blue component consisting of an OPF sensitive to blue color and a transparent readout circuit on another glass substrate for blue color imaging. Finally, we stack the fabricated components in layers. As a result, we obtain color images with 128 × 96 pixels through a shooting experiment of the stacked structure.

In this paper, a complementary metal oxide semiconductor (CMOS) image sensor for low-power operation with a variable frame rate is presented. The operating current of the CIS can be reduced by varying the frame rate of the CMOS image sensor (CIS), thus decreasing power consumption effectively. The proposed image sensor has two modes: normal mode and lowpower mode. In the low-power mode, the operating current of the pixel array can be controlled according to the frame rate of the CIS. At lower frame rates, the operating current of the pixel array is smaller than in the normal mode. The power consumption of the proposed CIS is determined by the on and off ratios of the bias current. The proposed image sensor was fabricated and measured with a 2-poly 4-metal 0.35 μm standard CMOS process.

We describe a lensless computational far-field imager that responds to thermal infrared light (8–14 μm) and comprises a spiral binary phase grating integrated with an 80 × 80-pixel microbolometer array followed by Fourier-domain computational image reconstruction. We believe this is the first hardware demonstration of computational diffractive imaging in thermal infrared.

Tone mapping algorithms have been utilized to adapt wide dynamic range (WDR) images to limited dynamic range (LDR) displays while still maintaining images’ details. This work presents an analog current mode circuitry approach to realize a tone mapping algorithm that includes both global and local operations in the image array periphery. Compared with the other in-pixel analog implementations, this work can keep the best pixel pitch and fill factor performance of WDR image sensors. In addition, by using joint global and local tone mapping operators, details of the images are preserved and the contrast of the image is also enhanced when compared with other hardware implementations realizing only one of these operators. In this work, a novel WDR analog divider is designed and presented to handle the WDR input pixel value. The circuit, designed in a TSMC 0.35 µm complementary metal oxide semiconductor (CMOS) technology, was simulated with thirty WDR images in order to evaluate our implementation. Peak signal to noise ratio (PSNR) and structural similarity (SSIM) scores were calculated to evaluate the hardware simulation and compared to a software implementation in MATLAB.

A gamma function is essential for adjusting the response of any display device. In the case of endoscopy, it is even more important because for endoscopy it is not only the satisfaction of the user but the diagnosis of patient’s problems that could lead to life and death decision sometimes. In this paper, a technique of approximating the gamma function is applied using a piecewise linear method. It has 10 bits input and output pixels of color (RBG) channels. VHDL is used to describe the function and implemented in a Spartan 6 FPGA to achieve high computation and parallel processing. It was tested on a small endoscopy camera called NanEye. The system has 31 reconfigurable gamma function values from 1 to 4 with 0.1 intervals. It has a small footprint in terms of memory and no specialized DSP processor. The average mean absolute error of the implemented solution is 2.1747. The system can process up to 750 million pixel components per second in a Spartan 6 FPGA.

Multi- and Hyperspectral Imaging (HSI) are characterized by the discrepancy between the dimensionality of hyperspectral image and video data and the dimensionality of the spectral detectors. This issue has been addressed by various schemes, including the Snapshot Mosaic Multispectral Imaging architecture, where each pixel (or group of pixels) is associated with a single spectral band. An unavoidable side effect of this design is the hard trade-off between spatial and spectral resolution. In this work, we propose a formal approach for overcoming this tradeoff by formulating the problem of full resolution recovery as a low rank Matrix Completion problem. Furthermore, we extend the traditional formulation of Matrix Completion by introducing non-negativity constraints during the recovery process, thus significantly enhancing the reconstruction quality. Experimental results suggest that the Non-Negative Matrix Completion (NN-MC) framework is capable of estimating a high spatial and spectral resolution hypercube from a single exposure, surpassing state-of- the-art schemes like the nearest-neighbors as well as the unconstrained Matrix Completion techniques.

Implementation of a motion detection algorithm in a very low power consuming image sensor is very constrained. A trade-off between the movement detection robustness and quantization level of pixel’s signals for a grayscale image, had to be established. Simulations have been made for both quantization resolution and frame rate. Obtained results will help us design an optimized ultralow power smart sensor.