Abstract: Fruit quality is a critical determinant of market value, consumer acceptance, and commercial competitiveness in the modern fruit industry. Accurate quality evaluation plays an essential role in ensuring agricultural product safety, enabling premium pricing for high-quality products, reducing postharvest losses, and improving overall industrial efficiency. Conventional fruit quality assessment methods were often destructive, labour-intensive, time-consuming, and unsuitable for rapid large-scale inspection. In recent decades, non-destructive detection technologies have emerged as important alternatives because of their rapid response, non-invasive operation, high-throughput capability, and potential for online integration. This review summarized recent advances in spectral- and imaging-based non-destructive detection technologies and related equipment for fruit quality evaluation. Eight representative technologies were comprehensively discussed, including visible/near-infrared (Vis/NIR) spectroscopy, hyperspectral imaging (HSI), X-ray imaging, laser speckle imaging, Raman spectroscopy, fluorescence spectroscopy, terahertz (THz) spectroscopy, and machine vision (MV). For each technology, the sensing principles, application status, technical characteristics, and developmental trends were critically analysed. Particular attention was given to their applications in internal quality assessment, external defect detection, physiological status monitoring, disease and pest identification, safety inspection, and intelligent grading. In addition to summarizing individual technologies, this review further compared their sensing capabilities, penetration depth, spatial resolution, engineering feasibility, suitable application scenarios, and online implementation potential. Different non-destructive detection technologies exhibited significant differences in information perception dimensions and target characteristics. Therefore, they were inherently complementary rather than mutually substitutive. From the perspective of intelligent fruit quality evaluation, future technological development would increasingly rely on multimodal sensing and collaborative information fusion to achieve more comprehensive, accurate, and robust quality assessment. Despite substantial progress, most existing studies were still conducted under controlled laboratory conditions, and several critical bottlenecks continued to restrict large-scale industrial application. One major challenge lay in the acquisition of high-quality detection signals. Another important limitation concerned the construction of robust predictive models. Furthermore, many existing models remained “black box” systems lacking mechanistic interpretability. In addition, although multimodal sensing had attracted increasing attention, efficient fusion of heterogeneous data from different sensing modalities remained insufficient because of differences in spatial scale, feature representation, and data structure. To address these challenges, future research should focus on several important directions. First, greater emphasis should be placed on elucidating the mechanisms underlying fruit quality formation and the interaction between biological tissues and detection signals, thereby establishing a stronger theoretical foundation for non-destructive sensing. Second, core sensing technologies should be further improved through enhanced signal acquisition strategies, high-sensitivity detectors, optimized optical configurations, and advanced interference suppression methods. Third, artificial intelligence and data-driven approaches should be deeply integrated into non-destructive detection systems. The development of interpretable, lightweight, transferable, and adaptive intelligent models based on deep learning, transfer learning, and multimodal data fusion would be essential for improving detection accuracy and robustness. Fourth, future equipment development should promote online, portable, miniaturized, and intelligent detection systems capable of stable operation in complex industrial environments. Finally, the establishment of standardized databases, unified evaluation criteria, and collaborative industrial application frameworks would be critical for accelerating technology transfer and large-scale commercialization. Overall, the integration of advanced sensing technologies, intelligent algorithms, and industrial equipment was expected to drive the next generation of non-destructive fruit quality detection systems and provide important technical support for intelligent postharvest management and modern smart agriculture.