Yajie Qin

An Auto-exposure algorithm for detecting high contrast lighting conditions

The proposed algorithm provides fast and accurate auto-exposure capability for digital still cameras. For normal lighting conditions, the number of preview frames and the exposure error are within 3.5 frames and 3.92%. For high contrast lighting conditions, the number of preview frames and the exposure error are within 8.8 frames and 6.56%. Furthermore, it offers accurate detection for both back lit and excessive front lit conditions at the same time and make proper exposure to the main object.

A 1.8-V 770-nW biopotential acquistion system for portable applications

In this paper, an ultra-low power biopotential acquisition system is presented. The proposed system contains a tunable band-pass amplifier, a VGA with strong capacitive driving capability and a successive approximation ADC. To make the gain and bandwidth programmable, a T-switch with huge cut-off resistor is utilized to eliminate the effect of low frequency doublets. An improved low kick-back noise low offset latch brings a significant power reduction to the SAR ADC. The design was implemented in 0.18 μ m CMOS process and consumes 770nW from a 1.8-V single supply. The analog front-end achieves a 193Hz to 407Hz programmable bandwidth, 41.5dB to 62dB tunable gain, 3.7μVrms (0.025Hz~200Hz) input referred noise, noise efficiency figure (NEF) of 3.1. The SAR ADC achieves 11.7-bit ENOB.

A wireless 8-channel ECG biopotential acquisition system for dry electrodes

A wireless 8-channel biopotential acquisition system for capturing electrocardiogram (ECG) using dry electrodes is presented. The ECG system consists of copper electrodes, a micropowered 8-channel custom ASIC, and an off-the-shelf microprocessor and bluetooth radio. Each analog channel of the custom ECG front-end is composed of a chopper-modulated instrumentation amplifier (CMIA) with chopping spike filter (CSF), a programmable gain amplifier (PGA), and a output buffer. Implemented in standard a 0.35 μm CMOS technology, the ECG front-end consumes 101 μA from a 2.7 V supply, occupying 5 mm2 of chip area. Measurement results show an input impedance of 1 G Ω, an input-referred noise of 0.97 μVrms (0.5 ~ 100 Hz), and a CMRR of 114 dB. Finally, a complete wireless 8-channel ECG monitoring system incorporating this analog front-end is demonstrated, showing successful recordings of a capture ECG waveform using a smart phone.

Design of an ultra-low power SAR ADC for biomedical applications

In this paper, an ultra-low power 12-bit 2kS/s successive approximation register (SAR) analog-to-digital converter (ADC) is presented. For power optimization, the voltage supply of the digital part is lowered, and the offset voltage of the latch is self-calibrated. Targeted for lower noise, a low kick-back noise latch is proposed. The chip was fabricated using 0.18µm 1P6M CMOS technology. The ADC achieves SNDR of 61.8dB and dissipates only 455nW, resulting in a figure of merit (FOM) of 220fJ/conversion-step. The ADC core occupies an active area of 674×639µm2.

System-on-chip for mega-pixel digital camera processor with auto control functions

A system on a chip design of digital camera processor for mega-pixel CCD digital camera is proposed. The processor integrates pipelined real-time digital image processor, JPEG codec and AE/AWB/AF controller. It supports 8-bit MCU and synchronous DRAM and provides interfaces to CCD, LCD, TV, PCMCIA, USB and EPP. The processor works in view-find, snapshot, playback and transfer modes. The snapshot speed is 5fps at 1.3M pixels and the view-find speed is fps. The chip is fabricated with 0.25 μm CMOS technology and has 200K logic gates with 12KB internal memory. The chip area is 25mm2 and the power consumption is 500mW at 54MHz system clock.A system on a chip design of digital camera processor for mega-pixel CCD digital camera is proposed. The processor integrates pipelined real-time digital image processor, JPEG codec and AE/AWB/AF controller. It supports 8-bit MCU and synchronous DRAM and provides interfaces to CCD, LCD, TV, PCMCIA, USB and EPP. The processor works in view-find, snapshot, playback and transfer modes. The snapshot speed is 5fps at 1.3M pixels and the view-find speed is fps. The chip is fabricated with 0.25 mm CMOS technology and has 200K logic gates with 12KB internal memory. The chip area is 25mm2 and the power consumption is 500mW at 54MHz system clock.

A 3.3-V, 1.9-GHz, high linear CMOS up-mixer with multi-tanh linearization technique

This paper presents a single side band up-mixer implemented in SMIC 0.35 mum CMOS technology. It can be used in low-IF direct conversion PCS 1900 (1850-1910MHz) transceiver systems. The mixer is based on multi-tanh linearization technique and achieves high linearity. It operates at a single power supply of 3.3V and consumes only 6mA. The up-mixer with the output buffer achieves an IIP3 of 8dBm and a 1-dB compression point of 0dBm

The application of EMD in activity recognition based on a single triaxial accelerometer

Activities recognition using a wearable device is a very popular research field. Among all wearable sensors, the accelerometer is one of the most common sensors due to its versatility and relative ease of use. This paper proposes a novel method for activity recognition based on a single accelerometer. To process the activity information from accelerometer data, two kinds of signal features are extracted. Firstly, five features including the mean, the standard deviation, the entropy, the energy and the correlation are calculated. Then a method called empirical mode decomposition (EMD) is used for the feature extraction since accelerometer data are non-linear and non-stationary. Several time series named intrinsic mode functions (IMFs) can be obtained after the EMD. Additional features will be added by computing the mean and standard deviation of first three IMFs. A classifier called Adaboost is adopted for the final activities recognition. In the experiments, a single sensor is separately positioned in the waist, left thigh, right ankle and right arm. Results show that the classification accuracy is 94.69%, 86.53%, 91.84% and 92.65%, respectively. These relatively high performances demonstrate that activities can be detected irrespective of the position by reducing problems such as the movement constrain and discomfort.

A low power analog front-end for portable biopotential acquisition systems

An ultra-low power analog front end is presented to capture ExG (including ECG, EEG and EMG) signals. It consists of a capacitively coupled chopper instrument amplifier, a chopping spike filter and a VGA. Implemented in standard 0.35μm CMOS technology, this analog front end occupying 0.54 mm2 chip area, while consuming only 3.96μA from 3.3 V supply. Measurement results show an input-referred noise of 0.85 μVrms (0.5-100 Hz), 115dB CMRR, 100dB PSRR, and a noise efficiency factor of 6.48.

Biofeedback neuromuscular electrical stimulation front-end for dysphagia treatment

A dedicated front-end for biofeedback neuromuscular electrical stimulation (NESM) system is proposed. For controllable dysphagia treatment, the integrated circuit (IC) provides a stimulator front-end with programmable stimulation parameters (2μA-1mA current amplitude, DC-2KHz frequency, and variable duty cycle) and an electromyogram/impedance (EMG/ETI) readout front-end with programmable gain and bandwidth (42-80dB, 0.1Hz-1.2KHz) for biofeedback. Area-efficient, low-power but high precision current controlling is achieved by inducing the sigma-delta modulator technique in the stimulation channel. The measured impedance and EMG signal are used to determine the stimulation parameters, enabling a closed loop optimized treatment. The proposed front-end is fabricated in a 0.18 μm standard CMOS process technology and dissipates a peak power of 2.3 mW at the supply voltage of 1.8 V. Measurement results on a live person are also provided to validate the systems effectiveness.

A Mobile-Based High Sensitivity On-Field Organophosphorus Compounds Detecting System for IoT-Based Food Safety Tracking

A mobile-based high sensitivity absorptiometer is presented to detect organophosphorus (OP) compounds for Internet-of-Things based food safety tracking. This instrument consists of a customized sensor front-end chip, LED-based light source, low power wireless link, and coin battery, along with a sample holder packaged in a recycled format. The sensor front-end integrates optical sensor, capacitive transimpedance amplifier, and a folded-reference pulse width modulator in a single chip fabricated in a 0.18u2009μm 1-poly 5-metal CMOS process and has input optical power dynamic range of 71u2009dB, sensitivity of 3.6u2009nW/cm2 (0.77u2009pA), and power consumption of 14.5u2009μW. Enabled by this high sensitivity sensor front-end chip, the proposed absorptiometer has a small size of 96u2009cm3, with features including on-field detection and wireless communication with a mobile. OP compound detection experiments of the handheld system demonstrate a limit of detection (LOD) of 0.4u2009μmol/L, comparable to that of a commercial spectrophotometer. Meanwhile, an android-based application (APP) is presented which makes the absorptiometer access to the Internet-of-Things (IoT).