Abstract: A system and method for visual simultaneous localization and mapping. The method including: extracting a blend of landmarks; associating descriptors and patches of pixels with the extracted landmarks; using the descriptors and patches of pixels, estimating a camera pose by performing feature matching and relative pose estimation; performing joint multi-objective pose optimization over photometric residuals and geometric residuals; updating a local map by performing Bundle Adjustment on the estimated pose; marginalizing extracted landmarks from the local map that are older than a predetermined number of keyframes and adding the descriptors associated with the marginalized landmarks to a global map; where there are loop closure candidates, performing point matching between a keyframe associated with the loop closure candidate and a keyframe most recently added to the global map; and rejecting the keyframe associated with the loop closure candidate if the number of matches is below a predetermined threshold.

Abstract: A pallet includes a deck having an upper surface for supporting goods thereon. The deck includes a plurality of ribs extending downward. According to one feature, the ribs may form plurality of stringers extending downward below the deck. Some of the ribs may be shorter in order to form a notch at each end of the stringers for engaging supports in a rack. In another feature, a plurality of locating features project upward from the surface of the deck. The plurality of locating features may each be generally in the shape of a pyramid for being received in the recessed bases of stackable plastic milk jugs.

Abstract: According to one aspect, there are systems and methods for generating data indicative of a three-dimensional representation of a scene. Current depth data indicative of a scene is generated using a sensor. Salient features are detected within a depth frame associated with the depth data, and these salient features are matched with a saliency likelihoods distribution. The saliency likelihoods distribution represents the scene, and is generated from previously-detected salient features. The pose of the sensor is estimated based upon the matching of detected salient features, and this estimated pose is refined based upon a volumetric representation of the scene. The volumetric representation of the scene is updated based upon the current depth data and estimated pose. A saliency likelihoods distribution representation is updated based on the salient features. Image data indicative of the scene may also be generated and used along with depth data.

Abstract: A pallet includes a deck having an upper surface for supporting goods thereon. The deck includes a plurality of ribs extending downward. According to one feature, the ribs may form plurality of stringers extending downward below the deck. Some of the ribs may be shorter in order to form a notch at each end of the stringers for engaging supports in a rack. In another feature, a plurality of locating features project upward from the surface of the deck. The plurality of locating features may each be generally in the shape of a pyramid for being received in the recessed bases of stackable plastic milk jugs.