Industrial Technology Research Institute

Novatek Microelectronics Corp.

Po-Wei Chen

Chun-Hung Chen

Hung-Cheng Kuo

Chieh-Cheng Chen

Shien-Chun Luo

Tsan-Huang Chen

Chi-Hung Wang

Yu-Hong Shen

Che-Jui Chang

A dynamic calibration method for heterogeneous sensors includes: sensing dynamic objects by a first sensor to generate first sensing data; sensing the dynamic objects by a second sensor to generate second sensing data; performing feature matching between the first sensing data and the second sensing data to determine first valid data and second valid data, and identifying a tracked object from the dynamic objects based on the first valid data and the second valid data; performing feature comparison between the first valid data and the second valid data corresponding to the tracked object to calculate data errors between the first sensor and the second sensor; and calculating a calibration parameter based on the first valid data and the second valid data when the number of the data errors exceeds an error threshold, and adjusting the first sensing data and the second sensing data based on the calibration parameter.

Dynamic calibration system and dynamic calibration method for heterogeneous sensors

A method for calculating state of health of battery, performed by a processing control circuit, includes: charging a battery under test to an upper voltage with a first constant current through a charging circuit and an electrical measuring circuit, discharging the battery under test to a lower voltage with a second constant current through the charging circuit and the electrical measuring circuit, obtaining an accumulated discharge capacity from the upper voltage to the lower voltage through the electrical measuring circuit, and calculating a state of health of the battery under test at least according to the accumulated discharge capacity and a design capacity.

Method and device for calculating state of health of battery

The disclosure provides an image processing device. An image processing device includes a neural network circuit, a gain control circuit, and an image merge circuit. The neural network circuit is configured to receive an input image, and perform an image processing operation on the input image according to fixed parameters to output a first intermediate image. The gain control circuit, coupled to the neural network circuit, is configured to receive the first intermediate image from the neural network circuit, and multiply the first intermediate image by at least one amplitude gain to output a second intermediate image. The image merge circuit, coupled to the gain control circuit, configured to receive the second intermediate image and the input image, and combine the second intermediate image and the input image to obtain an output image. In addition, a method for image enhancement is also provided.

Image processing device and a method for image enhancement

A convolutional neural network (CNN) operation accelerator comprising a first sub-accelerator and a second sub-accelerator is provided. The first sub-accelerator comprises I units of CNN processor cores, J units of element-wise & quantize processors, and K units of pool and nonlinear function processor. The second sub-accelerator comprises X units of CNN processor cores, Y units of first element-wise & quantize processors, and Z units of pool and nonlinear function processor. The above variables I&#x2dc;K, X&#x2dc;Z are all greater than 0, and at least one of the three relations, namely, 'I is different from X', 'J is different from Y', and &#x201c;K is different from Z&#x201d;, is satisfied. A to-be-performed CNN operation comprises a first partial CNN operation and a second partial CNN operation. The first sub-accelerator and the second sub-accelerator perform the first partial CNN operation and the second partial CNN operation, respectively.

Dynamic multi-mode CNN accelerator and operating methods

A noise detection method including the following steps is provided. During different time periods, plural sets of driving signals are respectively transmitted to driving lines of the touch panel to drive sensing lines of the touch panel to generate plural sets of sensing signals. The plural sets of sensing signals are respectively received and calculated to obtain plural sets of summation signals. One set of summation signals includes first summation signals, and another set of summation signals includes second summation signals. A part or all of the first summation signals is replaced by the second summation signals. A signal value of a combination of the first and the second summation signals is calculated to obtain a summation thereof. The summation of the signal value of the combination is smaller than a summation of a signal values of the first summation signals before recombination.

Growing community of inventors

Hsinchu County, Taiwan

Po-Wei Chen