Laser speckle detection and infrared imaging are non-contact, non-destructive testing methods based on different detection principles. Both can simultaneously detect internal defects in components under thermal loading conditions applied to the component's surface. This paper proposes an image enhancement method for laser speckle detection by fusing infrared images. The images from these two detection results are registered and fused, forming a novel composite detection and fusion technique that combines laser speckle detection with infrared imaging. Under thermal loading conditions, fusing the images from the two detection methods maximizes the retention of defect information in the fused image, effectively improving the defect recognition rate of the inspected components. Furthermore, this composite detection system can be applied to in-situ field testing, significantly enhancing the efficiency of routine component inspections.

Objective This paper proposes a hierarchical fusion image enhancement method and constructs an integrated composite detection system to improve the accuracy and stability of defect recognition. In the newly revised "China’s Non-Destructive Testing Technology Development Roadmap", laser shearography and infrared imaging non-destructive testing technologies have been identified as key emerging non-destructive testing methods requiring focused development. The deep integration of laser and infrared detection technologies represents a novel non-destructive testing approach that utilizes optical characteristics across different wavelength bands, enabling large-area and non-contact inspection.

Methods This paper proposes an image enhancement method that leverages the advantage of infrared thermography in real-time monitoring of temperature variations in objects. By aligning speckle images and infrared images to the same coordinate system through methods such as projective transformation, affine transformation, or homography matrix, the location of defects can be rapidly identified. The sequence image processing approach is adopted to analyze image data across the entire time series, thereby extracting more comprehensive defect information. Fourier phase analysis is performed on the sequence images to further eliminate background noise. Morphological operations are applied to the phase maps obtained after Fourier phase transformation to extract defect boundaries. Subsequently, by combining with the original shearography sequence image data, the defect fringe patterns with the highest quality for each defect are extracted. This enables phase enhancement of the shearography images, significantly improving the contrast of speckle fringe patterns and achieving the goal of image enhancement.

Results and Discussions Based on the aforementioned research, experimental testing was conducted on functional components, with the physical image of the functional component shown in Fig.11. Through extensive comparative experiments and verification, the optimal parameters for laser speckle detection, infrared imaging detection, and composite detection of the tested components were further determined, leading to corresponding research conclusions.The functional components were fabricated using a 2 mm-thick carbon fiber reinforced composite material as the substrate, with a 1mm-thick rubber patch adhered to the surface. The overall dimensions of the components were 150 mm × 160 mm. Two types of defect specimens were designed and manufactured, each containing two columns of defects. As shown in Fig.12, the left column represents tight-bond defects, created by embedding two layers of polytetrafluoroethylene (PTFE) film between the functional component and the carbon fiber board. The right column in Fig.12 represents hole-type defects, which were produced by machining flat-bottom holes with a depth of 0.5 mm into the carbon fiber board to simulate depression defects.Through extensive experiments, various parameters such as loading methods, shear angles, and excitation modes were tested and validated to determine the optimal detection parameters. Fig.13(a) shows the infrared imaging detection result, Fig.13(b) displays the speckle detection result, and Fig.13(c) presents the composite fusion detection result. Overall analysis of the detection effectiveness indicates that the fusion detection result integrates the advantages of both methods. The size of defects detected via the composite method closely matches the dimensions of the prefabricated defects, with intuitive, clear, and distinct image quality. This enhances both detection quality and effectiveness, achieving the expected objectives.

Conclusions Under thermal loading conditions, composite detection combining laser speckle and infrared imaging is performed on the component surface. By employing the laser speckle detection image enhancement method that fuses infrared images, defect information obtained from both techniques is maximally preserved in the fused image, leading to the following conclusions: 1) The composite fusion detection yields better results compared to laser speckle detection or infrared detection alone. The laser speckle and infrared imaging composite detection integrates information from both methods, complementing their respective strengths, which can effectively enhance the defect recognition rate in non-destructive testing of various materials. 2) The laser speckle detection image enhancement method incorporating infrared fusion effectively addresses issues such as blurred edges and low contrast of defects in the original infrared thermal images. It also mitigates problems related to the significant impact of environmental factors on laser speckle detection, which often leads to relatively poor image quality. 3) The composite detection technology combining laser speckle and infrared imaging, enhanced by the method of fusing infrared images into laser speckle detection, can be promoted as a multifunctional, non-contact, and large-area inspection approach for in-situ field testing applications involving novel multifunctional materials.