Motion Of Small Objects Research Articles

Photoreceptor processing improves salience facilitating small target detection in cluttered scenes

Target detection amidst clutter is a challenging task for both natural and artificial vision, yet one solved at the level of neurons in the 3rd optic ganglion of insects. These neurons are capable of responding to the motion of small objects, even against complex moving backgrounds. While the basic physiology has been investigated, little is known about how these cells are able to reject background motion while robustly responding to such small stimuli. By recording intracellularly from fly photoreceptors stimulated with natural image sequences containing a target viewed against a complex moving background, we show that the process of target detection begins at the earliest stages of vision. The temporal processing by photoreceptors alone, in the absence of any spatial interactions, improved the discrimination of targets (essentially a spatial task) by around 70%. This enhancement of target salience can be explained by elaborate models of photoreceptor temporal non-linear dynamics. The application of the functional principals outlined in this work could be utilized in areas such as robotics and surveillance, medical imaging, or astronomy, anywhere it is necessary to detect a small item from a cluttered surround.

MO-D-332-04: Implementation of Variable Temporal Filtering in a High-Resolution, Region-Of-Interest, High-Sensitivity, Micro-Angiographic Fluoroscope (HSMAF) Detector

Purpose: Temporal filtering with operator‐selectable weighting factor has been implemented in a LabVIEW‐based Graphical User Interface (GUI) for use with a new high‐resolution, High‐Sensitivity, Micro‐Angiographic Fluoroscope (HSMAF) in order to enhance visualization during neuro‐interventional procedures. Method and Materials: The detector which is based on a CsI(Tl) phosphor, a light image intensifier, and a fiber‐optic taper coupled to a charge‐coupleddevice(CCD) camera, provides the angiographer with real‐time display of high‐resolution region‐of‐interest (ROI) images. The high sensitivity of the HSMAF enables the detector to work in fast frame‐rate, low‐dose x‐ray procedures such as fluoroscopy and roadmapping which are subject to noise. The variable temporal filtering is able to reduce the imagenoise and improve visualization of details of small structures such as stent struts. The selectable weighting factor which can be set before or even during acquisition, offers a fine adjustment tool for the compromise between motion blurring and noise reduction. The temporal filter weighting‐factor can be varied between 1 and 12, typically; however, in extremely noisy cases and small object motion it can go up to a higher value. Results: Variable temporal filtering and its resulting noise reduction, have significantly improved image quality and the visualization capability of the HSMAF system. The variable weighting factor gave the flexibility to control the amount of noise smoothing, even during the fluoroscopic procedure. Despite the increased processing time needed for temporal filtering, the GUI can still maintain the 30 fps display rate during fluoroscopy and roadmapping. Conclusion: The ability of variable noise reduction during roadmapping and fluoroscopy with the HSMAF provides angiographers and interventionalists with a capability of enchanced visualization of small vessel and endovascular device details, such as stent‐struts, hence making diagnoses and image guided interventions more accurate. (Support from NIH Grants R01NS43924, R01EB002873 and Toshiba Medical Systems Corporation).

Motion and vision: why animals move their eyes.

Nearly all animals with good vision have a repertoire of eye movements. The majority show a pattern of stable fixations with fast saccades that shift the direction of gaze. These movements may be made by the eyes themselves, or the head, or in some insects the whole body. The main reason for keeping gaze still during fixations is the need to avoid the blur that results from the long response time of the photoreceptors. Blur begins to degrade the image at a retinal velocity of about 1 receptor acceptance angle per response time. Some insects (e.g. hoverflies) stabilise their gaze much more rigidly than this rule implies, and it is suggested that the need to see the motion of small objects against a background imposes even more stringent conditions on image motion. A third reason for preventing rotational image motion is to prevent contamination of the translational flow-field, by which a moving animal can judge its heading and the distances of objects. Some animals do let their eyes rotate smoothly, and these include some heteropod molluscs, mantis shrimps and jumping spiders, all of which have narrow linear retinae which scan across the surroundings. Hymenopteran insects also rotate during orientation flights at speeds of 100-200 degrees s-1. This is just consistent with a blur-free image, as are the scanning speeds of the animals with linear retinae.

Neural circuit tuning fly visual interneurons to motion of small objects. I. Dissection of the circuit by pharmacological and photoinactivation techniques.

1. Visual interneurons tuned to the motion of small objects are found in many animal species and are assumed to be the neuronal basis of figure-ground discrimination by relative motion. A well-examined example is the FD1-cell in the third visual neuropil of blowflies. This cell type responds best to motion of small objects. Motion of extended patterns elicits only small responses. As a neuronal mechanism that leads to such a response characteristic, it was proposed that the FD1-cell is inhibited by the two presumably GABAergic and, thus, inhibitory CH-cells, the VCH- and the DCH-cell. The CH-cells respond best to exactly that type of motion by which the activity of the FD1-cell is reduced. The hypothesis that the CH-cells inhibit the FD1-cell and, thus, mediate its selectivity to small moving objects was tested by ablating the CH-cells either pharmacologically or by photoinactivation. 2. After application of the gamma-aminobutyric acid (GABA) antagonist picrotoxinin, the FD1-cell responds more strongly to large-field than to small-field motion, i.e., it has lost its small-field selectivity. This suggests that the tuning of the FD1-cell to small moving objects relies on a GABAergic mechanism and, thus, most likely on the CH-cells. 3. The role of each CH-cell for small-field tuning was determined by inactivating them individually. They were injected with a fluorescent dye and then ablated by laser illumination. Only photoinactivation of the VCH-cell eliminated the specific selectivity of the FD1-cell for small-field motion. Ablation of the DCH-cell did not significantly change the response characteristic of the FD1-cell. This reveals the important role of the VCH-cells in mediating the characteristic sensitivity of the FD1-cell to motion of small objects. 4. The FD1-cell is most sensitive to motion of small objects in the ventral part of the ipsilateral visual field, whereas motion in the dorsal part influences the cell only weakly. This specific feature fits well to the sensitivity of the VCH-cell to ipsilateral motion that is most pronounced in the ventral part of the visual field. The spatial sensitivity distribution of the FD1-cell matches also the characteristics of figure-ground discrimination and fixation behavior.

Neural circuit tuning fly visual neurons to motion of small objects. II. Input organization of inhibitory circuit elements revealed by electrophysiological and optical recording techniques.

1. The FD1-cell in the visual system of the fly is an identified visual interneuron that is specifically tuned to motion of small objects. In the companion paper it was shown that this response property is mediated by one of the two CH-cells, the VCH-cell, that inhibits the FD1-cell by GABAergic synapses. Here the input organization of the two CH-cells is analyzed by both electrophysiological and optical recording techniques. 2. Both CH-cells are excited by front-to-back motion in the ipsilateral and by back-to-front motion in the contralateral visual field. They respond maximally to binocular rotatory motion about the vertical axis of the animal. The latter response is only slightly less than the sum of the corresponding monocular response components. The relative contribution of the ipsi-and contralateral eye to the binocular response varies considerably between flies. In extreme cases it is dominated by either the ipsi- or the contralateral eye. The two CH-cells are not equally sensitive along the vertical axis of the eye. The DCH-cell has its sensitivity maximum in the dorsal part, the VCH-cell in the ventral part of the visual field. 3. The CH-cells have two arborizations, a large one in the posterior part of the third visual neuropil, the lobula plate, and a smaller one in the ipsilateral ventrolateral brain. With the calcium-sensitive dye fura-2 as an activity marker, it is analyzed which of these branches of the CH-cells receive the ipsi- and contralateral motion input, respectively. During motion in the preferred direction within the ipsilateral visual field, calcium accumulates only in the CH-cells' main arborization in the lobula plate but not in their branches in the ventrolateral brain, indicating that the arborization in the lobula plate is postsynaptic to the ipsilateral input. In contrast, contralateral motion in the preferred direction leads to calcium accumulation in both arborizations, suggesting that both are postsynaptic to contralateral input elements. During preferred direction motion in the upper or lower part of the ipsilateral visual field, calcium accumulates in only dorsal or ventral branches of the CH-cells' arborization in the lobula plate, respectively, revealing that their ipsilateral motion input is organized retinotopically. Because this arborization, most likely, is also the main output terminal of the CH-cells, it is both pre- and postsynaptic. This specific neuronal design is discussed with respect to its consequences for the mechanism of tuning the FD1-cell to motion of small objects.

Motion-compensated wavelet transform coding for color video compression

A variable-block-size multiresolution motion compensation (MRMC) scheme in which the size of a block is adapted to its level in the wavelet pyramid is proposed. This scheme not only considerably reduces the searching and matching time but also provides a meaningful characterization of the intrinsic motion structure. The variable-block-size approach also avoids the drawback of the constant-size MRMC in describing small object motion activities. After wavelet decomposition, each scaled subframe tends to have different statistical properties. An adaptive truncation process is implemented, and a bit allocation scheme similar to that in transform coding is examined. Four variations of the proposed motion-compensated wavelet video compression system are identified, and it is shown that the coding approach has a superior performance in terms of the peak-to-peak signal-to-noise ratio as well as the subjective quality.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Colour inputs to motion and object vision in an insect

Self-motion detectors of the dragonfly ventral cord have large fields that are sensitive to whole-field one- way motion with little habituation to the motion. In contrast, object-motion detectors have large fields, but their responses to the motion of small objects within the field habituate more easily and some are non-directional. Both types are large neurons that can be used for electrophysiological recording for long periods and are previously described in their anatomy and their responses to black and white moving patterns (Oldberg 1986). When tested with a flash of controlled intensity and wavelength, the self-motion neurons have a spectral sensitivity similar to that of photoreceptors with a peak in the green near 500 nm, but when tested with a moving edge, they have a single peak near 560 nm. They behave towards a pattern of two colours as if they are colour blind. Similar results are known for the bee and some butterflies. Forward-looking object-motion neurons have a spectral sensitivity curve that is rather flat from 380 nm to 580 nm, some with a peak in the UV. When they are tested with a moving coloured pattern on a differently coloured background, there is no value of the foreground brightness, which gives a null response, however the brightness is adjusted. The optimum response to a coloured object appears when the background is green and the object is another colour; usually blue is the most effective. Discrimination of a small object is less effective when the object is green. These results suggest that colour vision is associated with object-vision; and that object-motion detectors are not colour blind, but do not necessarily discriminate colours.

On the neuronal basis of figure-ground discrimination by relative motion in the visual system of the fly

A fly can discriminate an object ("figure") from its background on the basis of motion information alone. This information processing task has been analysed, so far, mainly in behavioural studies but also in electrophysiological experiments (Reichardt et al., 1983). The present study represents a further attempt to bridge the gap between the behavioural and the neuronal level. It is based on behavioural and electrophysiological experiments as well as on computer simulations. The characteristic properties of figureground discrimination behaviour impose specific constraints on the spatial integration properties of the output cells of the underlying neuronal network, the heterolateral interactions in their input circuitry, as well as on the range of variability of their response. These constraints are derived partly from previous behavioural studies (Reichardt et al., 1983), partly, however, from behavioural response characteristics which have not been addressed explicitly so far. They are interpreted in terms of one of the alternative model circuits shown by Reichardt et al. (1983) to be sufficient to account for figure-ground discrimination. It will be demonstrated, however, that this can be done equally well by means of a further alternative model circuit. These constraints are used in the electrophysiological analysis for establishing visual interneurones as output elements of the neuronal network underlying figure-ground discrimination. In the behavioural experiments on figure-ground discrimination as well as on the optomotor course control the yaw torque generated by the tethered flying fly under visual stimulation was used as a measure for the strength and time course of the reaction. Therefore, it has initially been proposed that the three Horizontal Cells, which are regarded as the output elements of the neuronal network underlying the optomotor reaction (e.g. Hausen, 1981), might also control yaw torque generation in figure-ground discrimination (Reichardt et al., 1983). New behavioural data show, however, that the Horizontal Cells do not meet all the constraints imposed on the presumed output cells of the figure-ground discrimination network: (1) The Horizontal Cells are not sensitive enough to motion of small objects. (2) The heterolateral interactions within their input circuitry are not in accordance with the behavioural data (see also Reichardt et al., 1983). (3) The variability found in the time course of certain components of the yaw torque response to relative motion of figure and ground cannot be explained by their response characteristics. Hence, the Horizontal Cells cannot account for figure-ground discrimination on their own and additional output cells of the optic lobes with different functional properties are required to accomplish this task.

On the neuronal basis of figure-ground discrimination by relative motion in the visual system of the fly

A fly can discriminate an object (figure) from its background on the basis of motion information alone. This information processing task has been analysed, so far, mainly in behavioural studies but also in electrophysiological experiments (Reichardt et al., 1983). The present study represents a further attempt to bridge the gap between the behavioural and the neuronal level. It is based on behavioural and electrophysiological experiments as well as on computer simulations. The characteristic properties of figureground discrimination behaviour impose specific constraints on the spatial integration properties of the output cells of the underlying neuronal network, the heterolateral interactions in their input circuitry, as well as on the range of variability of their response. These constraints are derived partly from previous behavioural studies (Reichardt et al., 1983), partly, however, from behavioural response characteristics which have not been addressed explicitly so far. They are interpreted in terms of one of the alternative model circuits shown by Reichardt et al. (1983) to be sufficient to account for figure-ground discrimination. It will be demonstrated, however, that this can be done equally well by means of a further alternative model circuit. These constraints are used in the electrophysiological analysis for establishing visual interneurones as output elements of the neuronal network underlying figure-ground discrimination. In the behavioural experiments on figure-ground discrimination as well as on the optomotor course control the yaw torque generated by the tethered flying fly under visual stimulation was used as a measure for the strength and time course of the reaction. Therefore, it has initially been proposed that the three Horizontal Cells, which are regarded as the output elements of the neuronal network underlying the optomotor reaction (e.g. Hausen, 1981), might also control yaw torque generation in figure-ground discrimination (Reichardt et al., 1983). New behavioural data show, however, that the Horizontal Cells do not meet all the constraints imposed on the presumed output cells of the figure-ground discrimination network: (1) The Horizontal Cells are not sensitive enough to motion of small objects. (2) The heterolateral interactions within their input circuitry are not in accordance with the behavioural data (see also Reichardt et al., 1983). (3) The variability found in the time course of certain components of the yaw torque response to relative motion of figure and ground cannot be explained by their response characteristics. Hence, the Horizontal Cells cannot account for figure-ground discrimination on their own and additional output cells of the optic lobes with different functional properties are required to accomplish this task.

Visual discrimination between small objects and large textured backgrounds.

THE remarkable ability of certain higher order insect visual neurones to discriminate between movement of a small object or target and a large textured background has been clearly demonstrated by Palka1,2 and O'Shea and Fraser-Rowell3. In addition, movement of a large textured background such as a windmill pattern is inhibitory to detection of motion of a small target by this neurone, known as the descending contralateral movement detector (DCMD). The response of this identifiable neurone to motion of small objects is inhibited by rotation of the windmill pattern over a specific range of high spatial frequencies of the pattern, but is augmented by low spatial frequencies in the pattern. For this neurone we have quantitatively determined the spatial frequency at which the effect of the windmill pattern becomes inhibitory.