Abstract: A semiconductor device includes an image recognition device having a convolution arithmetic processing circuit. The convolution arithmetic processing circuit includes a coefficient register where coefficients of an integration coefficient table are set, a product calculation circuit that calculates products of an input image and the coefficients, a channel register where a channel number of the integration coefficient table is set, a channel selection circuit that selects an output destination of a cumulative addition arithmetic operation on the basis of the channel number, and a plurality of output registers that store a result of the cumulative addition arithmetic operation. The integration coefficient table is a table where a plurality of input coefficient tables are integrated and the integration coefficient table has a size of N×N. The product calculation circuit can calculate data of N×N all at once.

Abstract: A semiconductor device includes an image recognition device having a convolution arithmetic processing circuit. The convolution arithmetic processing circuit includes a coefficient register where coefficients of an integration coefficient table are set, a product calculation circuit that calculates products of an input image and the coefficients, a channel register where a channel number of the integration coefficient table is set, a channel selection circuit that selects an output destination of a cumulative addition arithmetic operation on the basis of the channel number, and a plurality of output registers that store a result of the cumulative addition arithmetic operation. The integration coefficient table is a table where a plurality of input coefficient tables are integrated and the integration coefficient table has a size of N×N. The product calculation circuit can calculate data of N×N all at once.

Abstract: A water droplet detection device has an image capturing unit and a water droplet detection unit. The image capturing unit has a photographic optical system that an area captures an image of a predetermined area. The water droplet detection unit sets an arbitrary attention point in the captured image, a plurality of first reference points inside an imaginary circle of a predetermined radius having the attention point as a center the imaginary circle, and a plurality of second reference points corresponding to the first reference points outside the imaginary circle. The water droplet detection unit detects edge information between the first reference points and second reference points, and assesses a circularity strength of the edge information to detect a water droplet attached to the photographic optical system. The water droplet detection device can be used with an image conversion unit to form a three-dimensional object detection device.

Abstract: An in-vehicle image processing device processes camera images to determine a parkable area and includes an overhead image generation section generating an overhead image relative to a road surface by converting the view point of an image of a nearby vehicle and the road surface taken by a vehicle-mounted camera; a tire recognition section recognizing the tires of the nearby vehicle; a three-dimensional object recognition section recognizing three-dimensional objects including the nearby vehicle; an area-under-vehicle recognition section calculating the ground contact position of the tire and the above-ground height from the ground contact position to the vehicle body and recognizing an area under the vehicle body on the road surface from the overhead image; and a parkable area recognition section for recognizing the area occupied by the vehicle body and the area under the vehicle body as a non-parking area and recognizing a parkable area on the road surface.

Abstract: An image recognition apparatus 100 includes a gradient feature computation unit 120 configured to calculate, from an image divided into a plurality of blocks, gradient feature values for each of the plurality of blocks, a combination pattern storage unit 160 configured to store a plurality of combination patterns of the gradient feature values, and a co-occurrence feature computation unit 131 configured to calculate a co-occurrence feature value in a plurality of blocks for each of the plurality of combination patterns. Further, image recognition apparatus 100 includes an arithmetic computation unit 132 configured to calculate an addition value by adding the co-occurrence feature value calculated for each of the plurality of blocks for each of the plurality of combination patterns, a statistical data generation unit 140 configured to generate statistical data from the addition value.

Abstract: An in-vehicle image processing device processes camera images to determine a parkable area and includes an overhead image generation section generating an overhead image relative to a road surface by converting the view point of an image of a nearby vehicle and the road surface taken by a vehicle-mounted camera; a tire recognition section recognizing the tires of the nearby vehicle; a three-dimensional object recognition section recognizing three-dimensional objects including the nearby vehicle; an area-under-vehicle recognition section calculating the ground contact position of the tire and the above-ground height from the ground contact position to the vehicle body and recognizing an area under the vehicle body on the road surface from the overhead image; and a parkable area recognition section for recognizing the area occupied by the vehicle body and the area under the vehicle body as a non-parking area and recognizing a parkable area on the road surface.

Abstract: An on-vehicle control device includes: an image acquiring unit that acquires an image from an image-capturing device that captures the image of surroundings of a vehicle; an overhead view image generating unit that generates an overhead view image from the image acquired by the image acquiring unit, the overhead view image being a plan view of the surroundings of the vehicle seen from above the vehicle; a white line recognizing unit that recognizes possible white lines on a road surface in the overhead view image; a white line information saving unit that saves information including positional information of the possible white lines recognized by the white line recognizing unit; a white line position predicting unit that predicts a position of a possible white line to be reached, based on information about a previous possible white line saved by the white line information saving unit and vehicle behavior information; and a white line judging unit that excludes, among the possible white lines recognized by the

Abstract: A three-dimensional object detection has an image capturing device, an image conversion unit, a three-dimensional object detection unit, a three-dimensional object assessment unit, first and second foreign matter detection units and a controller. The image capturing device captures images rearward of a vehicle. The three-dimensional object detection unit detects three-dimensional objects based on image information. The three-dimensional object assessment unit assesses whether or not a detected three-dimensional object is another vehicle. The foreign matter detection units detect whether or not foreign matter has adhered to a lens based on the change over time in luminance values for each predetermined pixel of the image capturing element and the change over time in the difference between an evaluation value and a reference value. The controller outputs control commands to the other means to suppress the assessment of foreign matter as another vehicle when foreign matter has been detected.

Abstract: An in-vehicle surrounding environment recognition device includes: a photographic unit that photographs a road surface around a vehicle and acquires a photographic image; an application execution unit that recognizes another vehicle on the basis of the photographic image, and detects a relative speed of the other vehicle with respect to the vehicle; a reflection determination unit that, on the basis of the photographic image, determines upon presence or absence of a reflection of a background object from the road surface; a warning control unit that controls output of a warning signal on the basis of the result of recognition of the other vehicle; and a warning prevention adjustment unit that suppresses output of the warning signal on the basis of the relative speed of the other vehicle, if it has been determined that there is the reflection of the background object from the road surface.

Abstract: An on-vehicle image processor includes an imager mounted on a vehicle, an image convertor and compositor that converts images imaged by the imager into overhead images, a white line candidate area detector that detects a white line candidate area estimated to constitute a parking frame, an endpoint position detector that detects positions of endpoints, a vehicle behavior measurer that measures a movement amount of the vehicle, an endpoint movement position predictor that predicts moved positions of the endpoints detected by endpoint position detector, an endpoint movement amount judger, an endpoint position storage, a distance calculator that judges a degree of proximity of the endpoints, an information output device, an endpoint position oblivion judger that deletes the position of the endpoints satisfying a predetermined condition, a parking frame detector that detects a parking frame from the image, an accelerator opening degree detector and a vehicle behavior controller.

Abstract: A turbidity degree calculating unit calculates a degree of the white turbidity U on a surface of the lens based on a brightness gradient g of an image I (x, y), and a lens cleaning control unit sets a lens cleaning mode for spraying at least one of the cleaning fluid and the compressed air, which is performed by the lens cleaning unit, based on the calculated turbidity degree U.

Abstract: A three-dimensional object detection device has an image capturing unit, an object detection unit, first and second edge intensity calculation units, a day/night assessment unit and a controller. The day/night assessment unit assess whether it is currently daytime or nighttime when detecting a three-dimensional object based on the captured images. Upon assesses it is daytime, edges of a subject are extracted from a first edge extraction area, including a horizon reference area, and a threshold value for detecting the three-dimensional object is set based on the intensity of the edges in the first edge extraction area. Upon assesses it is nighttime, the edges of a subject are extracted from a second edge extraction area, including a road edge reference area, and a threshold value for detecting the three-dimensional object is set based on the intensity of the edges that are extracted from the second edge extraction area.

Abstract: A three-dimensional object detection device has an image capturing unit, an object detection unit, a nighttime assessment unit, a luminance detection unit, a luminance peak detection unit and a controller. The image capturing unit captures images rearward of a vehicle. The object detection unit detects a presence of an object from the captured images. The nighttime assessment unit assesses if nighttime has fallen. The luminance detection unit detects a luminance of image areas from the captured image. The luminance peak detection unit detects a peak in the luminance having a luminance gradient that is greater than or equal to a predetermined reference value from among the detected peaks in the luminance as a target luminance peak. The controller controls detection of the object in an image area in which the target luminance peak is detected when an assessment has been made that nighttime has fallen by the nighttime assessment unit.

Abstract: A vehicle-mounted environment recognition device includes an imaging unit that acquires an image taken by an imaging device; an image self-diagnosis unit that diagnoses contamination of a lens of the imaging device with respect to the image acquired by the imaging unit; an application execution unit that executes an application selected from a predetermined application group; and a fail determination unit that determines, on the basis of a diagnosis result from the image self-diagnosis unit, whether the lens contamination state is in an endurance range of the application and, if within the endurance range, sets the operation of the application executed by the application execution unit to a suppression mode. If outside the endurance range, contamination removal control is implemented, or a fail determination is made for the application. Contamination removal and fail determination can be implemented utilizing parameter adjustment and hardware, or an image-based self-diagnosis can be conducted.

Abstract: An in-vehicle image recognizer effectively detects a moving object from an image even when a lens has grime. In a detection sensitivity adjustor (50) which adjusts detection sensitivity to be increased according to a white turbidity level (U), the detection sensitivity of a vehicle detector (70) (image recognition application execution unit), which detects the other vehicle (6) (moving object) existing in the surrounding area of a vehicle (5) with a predetermined detection sensitivity from the image obtained by an imaging unit (10) disposed in the vehicle (5) to observe the surrounding area of the vehicle (5) through a lens (12) and convert the light signal of the observed surrounding area of the vehicle (5) into an image signal, is corrected based on the attachment level M of the attached matter such as dirt or water drops to the lens (12), which is calculated by an attachment level calculator (26).

Abstract: An on-vehicle image processor includes an imager mounted on a vehicle, an image convertor and compositor that converts images imaged by the imager into overhead images, a white line candidate area detector that detects a white line candidate area estimated to constitute a parking frame, an endpoint position detector that detects positions of endpoints, a vehicle behavior measurer that measures a movement amount of the vehicle, an endpoint movement position predictor that predicts moved positions of the endpoints detected by endpoint position detector, an endpoint movement amount judger, an endpoint position storage, a distance calculator that judges a degree of proximity of the endpoints, an information output device, an endpoint position oblivion judger that deletes the position of the endpoints satisfying a predetermined condition, a parking frame detector that detects a parking frame from the image, an accelerator opening degree detector and a vehicle behavior controller.

Abstract: A vehicle-mounted environment recognition device includes an imaging unit that acquires an image taken by an imaging device; an image self-diagnosis unit that diagnoses contamination of a lens of the imaging device with respect to the image acquired by the imaging unit; an application execution unit that executes an application selected from a predetermined application group; and a fail determination unit that determines, on the basis of a diagnosis result from the image self-diagnosis unit, whether the lens contamination state is in an endurance range of the application and, if within the endurance range, sets the operation of the application executed by the application execution unit to a suppression mode. If outside the endurance range, contamination removal control is implemented, or a fail determination is made for the application. Contamination removal and fail determination can be implemented utilizing parameter adjustment and hardware, or an image-based self-diagnosis can be conducted.

Abstract: A three-dimensional object detection device has an image capturing unit, an object detection unit, a nighttime assessment unit, a luminance detection unit, a luminance peak detection unit and a controller. The image capturing unit captures images rearward of a vehicle. The object detection unit detects a presence of an object from the captured images. The nighttime assessment unit assesses if nighttime has fallen. The luminance detection unit detects a luminance of image areas from the captured image. The luminance peak detection unit detects a peak in the luminance having a luminance gradient that is greater than or equal to a predetermined reference value from among the detected peaks in the luminance as a target luminance peak. The controller controls detection of the object in an image area in which the target luminance peak is detected when an assessment has been made that nighttime has fallen by the nighttime assessment unit.

Abstract: A water droplet detection device has an image capturing unit and a water droplet detection unit. The image capturing unit has a photographic optical system that an area captures an image of a predetermined area. The water droplet detection unit sets an arbitrary attention point in the captured image, a plurality of first reference points inside an imaginary circle of a predetermined radius having the attention point as a center the imaginary circle, and a plurality of second reference points corresponding to the first reference points outside the imaginary circle. The water droplet detection unit detects edge information between the first reference points and second reference points, and assesses a circularity strength of the edge information to detect a water droplet attached to the photographic optical system. The water droplet detection device can be used with an image conversion unit to form a three-dimensional object detection device.

Abstract: A three-dimensional object detection has an image capturing device, an image conversion unit, a three-dimensional object detection unit, a three-dimensional object assessment unit, first and second foreign matter detection units and a controller. The image capturing device captures images rearward of a vehicle. The three-dimensional object detection unit detects three-dimensional objects based on image information. The three-dimensional object assessment unit assesses whether or not a detected three-dimensional object is another vehicle. The foreign matter detection units detect whether or not foreign matter has adhered to a lens based on the change over time in luminance values for each predetermined pixel of the image capturing element and the change over time in the difference between an evaluation value and a reference value. The controller outputs control commands to the other means to suppress the assessment of foreign matter as another vehicle when foreign matter has been detected.