The traditional layout of light source based on array will inevitably lead to uneven illumination and BER in the room. In order to improve the illumination uniformity and reliability of the system, it is necessary to optimize the layout of the light source reasonably. In this paper, a 4m × 4m × 3 m room is used as the model to design a ring light source layout model with a single-LED array and lamp belts. In this model, the distance between the inner lamps of the 6 x 6 LED array is 0.3 m. The number of surrounding lamp beads is 316 and the distance between the lamps is 0.05 m. The simulation results show that the system illumination is about 437.08 lx and the illumination uniformity is 93.9%. At the same time, the system error rate is about 2.8716×10-7.The annular light source layout model designed in this paper takes into account the uniformity of indoor illumination distribution and the reliability of communication simultaneously, which can meet the indoor lighting and communication at the same time and also can provide an optimization method for the layout of indoor visible light communication.

In order to measure the light environment of the ships navigation at night better, so as to lay the foundation for the evaluation of the light environment of the ships navigation at night and reduce the impact on the safety of the ship's navigation at night, a method of measuring the brightness of the sea light environment by a color CCD camera was proposed. Based on the principle of imaging, the relationship between brightness and camera parameters was deduced and the chromaticity calibration experiment was used to fit the relationship between the three-stimulus value and the standard RGB. The brightness calibration experiment fits out the unknown parameters in the formula and then the brightness measurement formula was determined. The port of Dalian Newport cruise terminal was chosen to verify the measurement formula. The results show that the method can measure the brightness of the light environment of the ships navigation at night, which has high precision.

In view of the issue that in the process of locomotive motor commutator maintenance and processing, laser modulation and positioning type mica slot engraving system are widely used in real applications. However, it has low precision and requires a lot of human intervention. To overcome this problem, this paper proposed a precise positioning method of mica slot based on machine vision, which can accurately extract the edge of the motor commutator. The system first independently designs the precision compensation algorithm and constructs the positioning error correction model. Based on this, the embedded system platform is built to realize the automatic, rapid and accurate positioning the center line of the mica slot. Then the system achieves accurate calculation the deviation between the knife and the center line and control the bit tool to move to the correct engraved position. Experimental results indicate that the instrument can accurately calculate the position of the mica slot center line. Through the servo motor control burin to adjustment and aim the midline, the knife positioning error has been controlled between the positive and negative 0.02 mm, achieving the entire operation process automation and precision.

Diabetic macular edema (DME) is one of the important reasons leading to blindness. Its pathological features are mainly manifested in the accumulation of fluid in the retina. A method for segmentation of diabetic macular edema in optical coherence tomography (OCT) retinal images is proposed. Firstly, through the image preprocessing, we exclude the impact of speckle noise and blood vessels on the final segmentation results. We used the improved level set method to solve the problem of segmentation effectively and calculated the area of edema area, which provides quantitative analytic tools for clinical diagnosis and therapy. Finally, we validated the method in this study on 15 OCT retina images with DME adults. The precision, sensitivity and dice similarity coefficient (DSC) for DME segmentation are 81.12%, 86.90% and 80.05%, respectively.

In this paper, the FRT (fringe reflection technique) is used for the off-axis aspheric surface measurement during the manufacture stage of the beginning of polish, duo to its advantage of large dynamic range and high sensitivity. The measurement system coordinate and ray trace model are build using the laser tracker, and the calibration results of the camera calibration and screen calibration are introduced into the Zemax model and the ideal screen pixel point position can be got by ray tracing. The measurement of screen pixel point position is obtained by phase-shifting technique. The slope error of the surface is calculated and final results are got by integration. The measurement results of a SiC off-axis mirror obtained by the presented method and the CMM are compared and feasibility is verified. This method can be used to guide the manufacture of off-axis aspheric surface during the beginning of the polish.

Convolutional neural networks have recently been shown to have the highest accuracy for single image super-resolution (SISR) reconstruction. Most of the network structures suffer from low training and reconstruction speed, and still have the problem that one model can only be rebuilt for a single scale. For these problems, a deep cascaded network (DCN) is designed to reconstruct the image step by step. L2 and the perception loss function are used to optimize the network together, and then a high quality reconstructed image will be obtained under the joint action of each cascade. In addition, our network can get reconstructions of different scales, such as 1.5′, 2′, 2.5′, 3′, 3.5′ and 4′. Extensive experiments on several of the largest benchmark datasets demonstrate that the proposed approach performs better than existing methods in terms of accuracy and visual improvement.

In this paper, in view of the on-line inspection of the thickness of steel plate and its surface defects, a quality inspection system for appearance of steel is designed. The moire image on the surface of the steel plate is collected by a linear CCD camera. The fringe image is transformed by wavelet transform, and the phase information of wavelet transform coefficients corresponding to the wavelet ridge is extracted to reconstruct the 3D profile. Experimental results show that the thickness measurement precision of the steel plate is 0.08 mm and the measurement accuracy of the surface defect is 0.2 mm. The online detection speed is 6 m/s. The on-line detection of thickness and surface defect of steel plate can be realized.

Due to the interference such as sea waves, ships and light, it is difficult to accurately detect the sea-sky-line of the visible light maritime image. To improve the detection accuracy and robustness, a sea-sky-line detection method based on local Otsu segmentation and Hough transform is proposed. Firstly, high-frequency noise such as light spot in the gray image is rapidly suppressed by longitudinal median filter. Then, according to the image features, the gray image is divided into image blocks in longitudinal to compensate for inhomogeneity of illumination and limit the interference scope of ships to some image blocks. Afterwards, local Otsu segmentation is performed on the gray image to obtain the binary image where edge pixels are extracted, which suppresses the interference of waves. Finally, Hough transform is used to fit edge pixels to complete the sea-sky-line detection. Experimental re- sults show that the proposed method is relatively accurate, robust and real-time. The detection accuracy of the proposed method is 93.0%, which is significantly higher than that of three representative sea-sky-line detection methods.

In order to solve the problem of low sampling rate and to meet the requirements of high frequency swing mirror measurement, based on the principle of equivalent sampling, a multi time series high accuracy non-contact swing mirror detection system is designed. First, with the laser pulse control circuit, multi time sequence flashing laser lighting is achieved and the spot position is obtained on the CCD target plane. Then, according to the image of the laser spot position, calculate the mirror swing angle and angular velocity by the time interval between adjacent spot positions. The experimental results show that the angle resolution of the detection system is 0.005°. The time resolution can reach the microsecond order, and the angular velocity measurement error is less than ±7%. The system improves the sampling frequency of swing mirror measuring and meets the requirements of high frequency swing mirror test.

A high dynamic range image is generated by using multiple low dynamic range images with different exposure times. After the generation, the image noise will be further amplified, resulting in a severe degradation in the visual quality of the final high dynamic range image. In view of the problem that the generated image needs to retain the detail information of the high-lighted area in the low-exposure image and the detail information of the low-dark area in the high-exposure image and the image noise is related to the luminance, a noise suppression algorithm based on luminance partitioning and noise level estimation in the process of high dynamic range image fusion is proposed in this paper. Firstly, according to the luminance information of the image, different luminance regions of the low dynamic range image are determined. And then the overlapped blocks are used to estimate the noise level of the different luminance regions of the image, so as to guide the sparse denoising of the image. Finally, a high dynamic range image is generated by the processed low dynamic range images. The experimental results show that the proposed algorithm can effectively suppress noise, and the generated high dynamic range image has better visual quality.

The state-of-art camera calibration method requires the user to provide accurate pixel coordinates of calibration plate feature points. For some cameras with special sensing range, general calibration objects’ (such as calibration plates with a centimeter-long dimension) using range is outside their clear sensing range. Using these cameras to take a picture for general calibration objects, you can only get out-of-focused blurred images that can not accurately extract feature points’ pixel coordinates. This paper analyzes the influence on the phase of the structured light based on sine grating (abbreviated as sinusoidal structured light) when optical system is in defocus state. Based on the fact that the state of focus is independent of the phase of sinusoidal structured light, a method of phase-shifted sinusoidal structured light encoding by phase shift is proposed to encode the feature points on the calibration object and this method realizes the calibration of the camera under out-of-focus condition. The experimental results show that the maximal deviation of focal length from the real value is 0.47% and the maximal pixel reprojection error is 0.17 pixels. This paper provides a solution to camera calibration with a special sensing range.