{"id":3694,"date":"2026-01-27T03:56:43","date_gmt":"2026-01-27T03:56:43","guid":{"rendered":"https:\/\/ni-mech.com\/?p=3694"},"modified":"2026-01-27T04:00:07","modified_gmt":"2026-01-27T04:00:07","slug":"the-health-check-report-of-vision-modules","status":"publish","type":"post","link":"https:\/\/ni-mech.com\/ja\/the-health-check-report-of-vision-modules\/","title":{"rendered":"\u30d3\u30b8\u30e7\u30f3\u30e2\u30b8\u30e5\u30fc\u30eb\u306e\u300c\u5065\u5eb7\u8a3a\u65ad\u30ec\u30dd\u30fc\u30c8"},"content":{"rendered":"

Understanding MTF and SFR Testing in One Article<\/h2>\n

Inside the darkroom of a testing laboratory, a smartphone camera faces a carefully illuminated slanted-edge test chart. An engineer clicks the analysis button, and a smooth curve instantly appears on the screen\u2014descending from high to low\u2014accurately predicting how clearly this camera will capture the world.<\/p>\n

When evaluating a camera, a lens, or even an entire imaging system, sharpness<\/strong> is often the most intuitive requirement. But how can this subjective visual perception be transformed into objective, repeatable, and scientific data<\/strong>?<\/p>\n

This is precisely the role of Modulation Transfer Function (MTF)<\/strong> and Spatial Frequency Response (SFR)<\/strong>. They are the universal language<\/em> of optics and imaging, and the gold standard<\/strong> for quantitatively evaluating image sharpness and detail reproduction in vision modules.<\/p>\n

\n

01 Core Concepts: From Blur to Quantification<\/h2>\n

What Is MTF?<\/h3>\n

Simply put, MTF measures the \u201cfidelity\u201d of an imaging system<\/strong>. It describes how well the system can reproduce the contrast of black-and-white line patterns at different levels of fineness (spatial frequencies).<\/p>\n

The basic modulation formula is:<\/p>\n

M = (Imax \u2212 Imin) \/ (Imax + Imin)<\/strong><\/p>\n

The MTF value is defined as the ratio between the image modulation and the original object modulation.
An ideal MTF value of 1<\/strong> represents perfect reproduction.<\/p>\n

In testing, spatial frequency is measured in line pairs per millimeter (lp\/mm)<\/strong> and can be understood as the fineness<\/em> of image details. Higher spatial frequency corresponds to denser lines and finer details.<\/p>\n

A general rule applies: as spatial frequency increases, MTF typically decreases<\/strong>, because an imaging system\u2019s ability to reproduce extremely fine details is always limited.<\/p>\n

For example:<\/p>\n

    \n
  • \n

    At 20 lp\/mm<\/strong>, an MTF value above 80%<\/strong> indicates excellent image quality<\/p>\n<\/li>\n

  • \n

    If the MTF falls below 30%<\/strong>, image quality is already poor even at standard print sizes<\/p>\n<\/li>\n<\/ul>\n

    \n

    What Is SFR?<\/h3>\n

    SFR is often described as the \u201cdigital cousin\u201d of MTF<\/strong>. It is specifically designed to evaluate complete digital imaging systems<\/strong>, including the lens, image sensor, and image signal processing (ISP)<\/strong>.<\/p>\n

    The goal of SFR is to convert human-perceived \u201csharpness\u201d into measurable physical parameters.<\/p>\n

    SFR works by analyzing the imaging system\u2019s response to a slanted sharp edge<\/strong>, from which the system\u2019s overall frequency response can be derived.<\/p>\n

    \n

    02 Physical Principles: How Details Are Lost<\/h2>\n

    Why do images become blurred? From a physical perspective, image degradation in digital imaging systems results from multiple factors, which can be grouped into three main categories:<\/p>\n

    Hardware Limitations<\/h3>\n
      \n
    • \n

      Lenses inherently suffer from residual geometric aberrations and diffraction<\/p>\n<\/li>\n

    • \n

      Sensor pixel aperture and electrical crosstalk between pixels lead to detail loss<\/p>\n<\/li>\n<\/ul>\n

      Software Algorithm Effects<\/h3>\n

      To produce more visually pleasing images, algorithms such as noise reduction, beautification, and background blur actively modify pixel data. While reducing noise, they often remove genuine fine details<\/strong>.<\/p>\n

      Environmental Influences<\/h3>\n

      Camera shake, atmospheric turbulence, and stray light under strong illumination can all directly degrade image sharpness.<\/p>\n

      MTF and SFR quantify the combined impact of these factors by treating the imaging process as a signal-processing system<\/strong>:<\/p>\n

        \n
      • \n

        Input: original scene information<\/p>\n<\/li>\n

      • \n

        Output: final image<\/p>\n<\/li>\n<\/ul>\n

        The system modifies the contrast (amplitude) of different frequency components.
        The MTF\/SFR curve visually shows how much contrast is preserved at each spatial frequency<\/strong>.<\/p>\n

        A key principle is that the overall system MTF curve is the product of the MTF curves of individual components<\/strong> (lens, sensor, optical low-pass filter, etc.).
        Therefore, analyzing the MTF curve allows engineers to pinpoint bottlenecks<\/strong> and optimize system design with precision.<\/p>\n

        \n

        03 Testing in Practice: How Is Sharpness \u201cScored\u201d?<\/h2>\n

        Standardized testing methods are the foundation of result comparability.
        The internationally recognized standard for SFR testing is ISO 12233<\/strong>, first released in 2000 and continuously updated since.<\/p>\n

        Core Testing Principle<\/h3>\n

        The standard uses a slanted edge<\/strong> (typically tilted by ~5\u00b0) as the test target.
        This slight tilt is crucial\u2014it creates multiple phase samples, enabling oversampling<\/strong>.<\/p>\n

        This approach compensates for uncertainties caused by the discrete pixel array of image sensors, allowing accurate calculation of the Edge Spread Function (ESF)<\/strong>.<\/p>\n

        \n

        Standard Test Workflow<\/h3>\n
          \n
        1. \n

          Environment Setup<\/strong>
          Testing is performed in a darkroom using standardized illumination (e.g., 6500K color temperature, 800 lux), ensuring uniform lighting of the test chart.<\/p>\n<\/li>\n

        2. \n

          Image Capture<\/strong>
          The device under test (DUT) is securely mounted, precisely focused, and used to capture images of the test chart.<\/p>\n<\/li>\n

        3. \n

          Software Analysis<\/strong>
          Professional software such as Imatest Master<\/strong> or RIQA<\/strong> automatically detects the slanted-edge region and performs a sequence of calculations:<\/p>\n

            \n
          • \n

            Fit a supersampled Edge Spread Function (ESF)<\/strong><\/p>\n<\/li>\n

          • \n

            Compute the first derivative of ESF to obtain the Line Spread Function (LSF)<\/strong><\/p>\n<\/li>\n

          • \n

            Apply Fourier transform to the LSF and take the magnitude, resulting in the final SFR (MTF) curve<\/strong><\/p>\n<\/li>\n<\/ul>\n<\/li>\n<\/ol>\n

            \n

            Table 1: Key SFR Metrics Explained<\/h3>\n
            \n
            \n\n\n\n\n\n\n\n
            Metric<\/th>\nDefinition & Physical Meaning<\/th>\nTypical Applications<\/th>\n<\/tr>\n<\/thead>\n
            MTF50<\/strong><\/td>\nSpatial frequency at which MTF drops to 50%<\/td>\nHighly correlated with human-perceived sharpness; core metric for overall image sharpness<\/td>\n<\/tr>\n
            MTF50P<\/strong><\/td>\nFrequency where MTF drops to 50% of its peak value<\/td>\nEliminates low-frequency contrast loss (e.g., vignetting) to better reflect mid-to-high frequency detail<\/td>\n<\/tr>\n
            MTF30 \/ MTF10<\/strong><\/td>\nFrequencies where MTF drops to 30% or 10%<\/td>\nEvaluates extreme low-contrast detail resolution; important for high-contrast scenes such as text recognition<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<\/div>\n
            \n

            04 Practical Setup: Building Your Test System<\/h2>\n

            Depending on testing objectives, accuracy requirements, and budget, system configurations can be categorized into several levels.<\/p>\n

            Basic R&D and Validation Setup<\/h3>\n
              \n
            • \n

              ISO-compliant SFR test charts<\/strong> (slanted-edge, SFRplus, etc.)<\/p>\n<\/li>\n

            • \n

              Standard light sources<\/strong> with adjustable color temperature and illuminance<\/p>\n<\/li>\n

            • \n

              Image quality analysis software<\/strong> such as Imatest Master or RIQA<\/p>\n<\/li>\n<\/ul>\n

              \n

              Professional Optical Laboratory Setup<\/h3>\n
                \n
              • \n

                Addition of a collimator<\/strong>, which simulates infinity-focus conditions within a limited space<\/p>\n<\/li>\n

              • \n

                Essential for long-focus lenses and infinity-focus performance evaluation<\/p>\n<\/li>\n<\/ul>\n

                \n

                Advanced Automated Test Systems<\/h3>\n

                High-end labs deploy fully automated systems integrating:<\/p>\n

                  \n
                • \n

                  Motorized collimators<\/p>\n<\/li>\n

                • \n

                  High-precision six-axis motion platforms<\/p>\n<\/li>\n

                • \n

                  Automated image capture, multi-angle alignment, data analysis, and report generation<\/p>\n<\/li>\n<\/ul>\n

                  \n

                  Table 2: Typical Test Equipment Configurations<\/h3>\n
                  \n
                  \n\n\n\n\n\n\n\n\n\n
                  Equipment<\/th>\nCore Function & Requirements<\/th>\nRepresentative Models \/ Standards<\/th>\n<\/tr>\n<\/thead>\n
                  Test Charts<\/td>\nStandardized patterns with known reflectance<\/td>\nISO 12233 slanted edge, SFRplus, Siemens star<\/td>\n<\/tr>\n
                  Standard Light Source<\/td>\nStable, uniform, adjustable CCT and illuminance<\/td>\nLS-CCXL series (2300K\u201310000K)<\/td>\n<\/tr>\n
                  Analysis Software<\/td>\nROI detection, ESF\u2192LSF\u2192MTF computation, reporting<\/td>\nImatest Master, RIQA<\/td>\n<\/tr>\n
                  Collimator<\/td>\nSimulates infinity targets<\/td>\nIntegrated in RFT systems<\/td>\n<\/tr>\n
                  Integrated Test System<\/td>\nFully automated optical\/mechanical\/electrical testing<\/td>\nIS-RFT (supports up to 210\u00b0 FOV, SFR\/distortion\/CA testing)<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<\/div>\n
                  \n

                  Production-Line Quality Control<\/h3>\n

                  In mass production, priorities shift to speed, consistency, and automation<\/strong>. Integrated inline test equipment can complete image capture, analysis, and pass\/fail judgment within seconds, automatically uploading results to MES systems for real-time quality monitoring and traceability<\/strong>.<\/p>\n

                  \n

                  When engineers debate optimization strategies while studying overlapping MTF curves, they are not merely discussing abstract percentages.
                  Each improvement in the curve means:<\/p>\n

                    \n
                  • \n

                    Autonomous vehicles detect road signs earlier<\/p>\n<\/li>\n

                  • \n

                    Medical endoscopes reveal finer tissue details<\/p>\n<\/li>\n

                  • \n

                    Smartphones capture fleeting moments with greater clarity<\/p>\n<\/li>\n<\/ul>\n

                    Behind every MTF curve is a clearer world.<\/strong><\/p>","protected":false},"excerpt":{"rendered":"

                    Understanding MTF and SFR Testing in One Article Inside the […]<\/p>","protected":false},"author":2,"featured_media":3651,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[82],"tags":[],"class_list":["post-3694","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tips-tricks"],"acf":[],"_links":{"self":[{"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/posts\/3694","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/comments?post=3694"}],"version-history":[{"count":2,"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/posts\/3694\/revisions"}],"predecessor-version":[{"id":3696,"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/posts\/3694\/revisions\/3696"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/media\/3651"}],"wp:attachment":[{"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/media?parent=3694"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/categories?post=3694"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ni-mech.com\/ja\/wp-json\/wp\/v2\/tags?post=3694"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}